Digital PDFs
Documents
Guest
Register
Log In
AA-Z423B-TE
2000
142 pages
Original
8.9MB
view
download
OCR Version
14MB
view
download
Document:
MONITOR
Order Number:
AA-Z423B-TE
Revision:
000
Pages:
142
Original Filename:
aa-z423b-te.pdf
OCR Text
l MONITOR Contents PREFACE ix NEW AND CHANGED FEATURES xi FORMAT MON-1 MONITOR COMMANDS MON-3 DESCRIPTION MON-3 1 CLASS TYPES MON-4 2 CLASS-NAME QUALIFIERS MON-4 3 OUTPUTS MON-6 3.1 Display Output MON-6 3.2 Display Data MON-7 3.3 Screen Formats MON-7 3.4 3.3.1 Single-Statistic Screen for System Classes MON-8 3.3.2 Multiple-Statistic Screen for System Classes MON-8 3.33 Component-Classes Screen MON-9 Recording File Output MON-10 3.4.1 Disk Space for Recording Files MON-11 342 Recording File Version Compatibility MON-11 3.5 Summary Output MON-11 3.6 Multifile Summaries MON-12 4 3.6.1 Interpreting Multifile Summary Reports MON-12 362 Using Multifile Summary Reports in Single-Node Environments MON-14 3.6.3 Using Multifile Summary Reports in VAXcluster Environments ERROR MESSAGES MON-15 MON-15 MONITOR Contents COMMAND QUALIFIERS MON-17 /BEGINNING MON-18 /BY_NODE MON-19 JCOMMENT MON-20 /DISPLAY MON-21 /ENDING MON-22 /FLUSH_INTERVAL MON-23 /INPUT MON-24 /INTERVAL MON-25 /NODE MON-27 JOUTPUT MON-28 /RECORD MON-29 /SUMMARY MON-30 /VIEWING_TIME MON-31 COMMANDS MON-32 CONVERT MON-33 EXECUTE (@) MON-34 EXIT MON-35 HELP MON-36 INITIALIZE MON-37 MONITOR ALL_CLASSES MON-38 MONITOR CLUSTER MON-40 MONITOR DECNET MON-45 MONITOR DISK MON-47 MONITOR DLOCK MON-50 MONITOR FCP MON-52 MONITOR FILE_SYSTEM_CACHE MON-54 MONITOR 10 MON-58 MONITOR LOCK MON-60 MONITOR MODES MON-62 MONITOR Contents MONITOR PAGE MON-65 MONITOR POOL MON-67 MONITOR PROCESSES MON-69 MONITOR SCS MON-72 MONITOR STATES MON-76 MONITOR SYSTEM MON-79 SET DEFAULT MON-82 SHOW DEFAULT MON-83 EXAMPLES SECTION A SUPPLEMENTAL MONITOR INFORMATION MON-84 MON-91 A1 THE MONITOR RECORDING FILE MON-91 A2 CONVENTIONS MON-92 A3 DIGITAL CONTROL RECORDS MON-93 A31 File Header Record MON-93 A3.2 System Information Record MON-95 A33 Node Transition Record MON-96 A4 CLASS RECORDS MON-97 A4 Class Type Formats MON-97 A4 Class Header MON-98 A4.1.2 Class Prefix (Component Classes Only) MON-99 A4.13 Data Block MON-100 A42 PROCESSES Class Record MON-100 A43 STATES Class Record MON-102 A44 MODES Class Record MON-104 A45 PAGE Class Record MON-106 A4.6 1/0 Class Record MON-108 AA47 FCP Class Record MON-109 A.4.8 POOL Class Record MON-111 A49 LOCK Class Record MON-112 v MONITOR Contents A.4.10 DECNET Class Record MON-113 A.4.11 FILE_SYSTEM_CACHE Class Record MON-114 A4.12 DISK Class Record MON-116 A4.13 DLOCK Class Record MON-118 A.4.14 SCS Class Record MON-120 A.4.15 SYSTEM Class Record MON-122 A4.16 CLUSTER Class Record MON-124 —_— INDEX EXAMPLES MON-1 Single-Statistic Screen MON-8 MON-2 Sample Muitiple-Statistic Screen MON-9 MON-3 Sample Multiple-Statistic Screen (Data Expressed as Percentages) MON-9 MON-4 Sample Component Statistics Screens MON-10 MON-5 Sample Single-Node Multifile Summary MON-14 MON-6 Sample VAXcluster Multifile Summary MON-16 - FIGURES MON-1 File Header Record Format MON-94 MON-2 System Information Record Format MON-96 MON-3 Node Transition Record Format MON-97 MON-—4 Class Record Format MON-98 MON-5 Class Header Format MON-98 MON-6 Class Prefix Format MON-99 MON-7 PROCESSES Class Record Format MON-101 MON-8 STATES Class Record Format MON-103 MON-9 MODES Class Record Format MON-105 MON-10 PAGE Class Record Format MON-107 MON-11 1/0 Class Record Format MON-108 MONITOR Contents MON-12 FCP Class Record Format MON-110 MON-13 POOL Class Record Format MON-111 MON-14 LOCK Class Record Format MON-113 MON-15 DECNET Class Record Format MON-114 MON-16 FILE_SYSTEM_CACHE Class Record Format MON-115 MON-17 DISK Class Record Format MON-117 MON-18 DLOCK Class Record Format MON-119 MON-19 SCS Class Record Format MON-121 MON-20 SYSTEM Class Record Format MON-123 MON-21 CLUSTER Class Record Format MON-124 TABLES MON-1 MONITOR Class-Name Qualifiers MON-2 Relationship of MONITOR Command Qualifier to MON-4 Event MON-26 MON-3 MONITOR Record Types MON-92 MON-4 Descriptions of File Header Record Fields MON-95 MON-5 Descriptions of System Information Record Fields MON-96 MON-6 Descriptions of Node Transition Record Fields MON-97 MON-7 Descriptions of Class Header Fields MON-99 MON-8 Descriptions of Class Prefix Fields MON-100 MON-9 Descriptions of PROCESSES Class Record Fields MON-102 MON-10 Descriptions of STATES Class Record Fields MON-103 MON-11 Descriptions of MODES Class Record Fields MON-105 MON-12 Descriptions of PAGE Class Record Fields MON-107 MON-13 Descriptions of 1/O Class Record Fields MON-109 MON-14 Descriptions of FCP Class Record Fields MON-110 MON-15 Descriptions of POOL Class Record Fields MON-112 MON-16 Descriptions of LOCK Record Fields MON-113 MON-17 Descriptions of DECNET Class Record Fields MON-114 MON-18 Descriptions of FILE_SYSTEM_CACHE Class Record Fields MON-115 MONITOR Contents viii MON-19 Descriptions of DISK Class Record Fields MON-118 MON-20 Descriptions of DLOCK Class Record Fields MON-119 MON-21 Descriptions of SCS Class Record Fields MON-122 MON-22 Descriptions of SYSTEM Class Record Fields MON-123 MON-23 Descriptions of CLUSTER Class Record Fields MON-125 Preface Intended Audience This manual is intended for VAX/VMS system managers, operators, and system programmers. Structure of This Document This document is composed of six major sections. The Format Section is an overview of MONITOR and is intended as a quick reference guide. The format summary describes the DCL command that invokes MONITOR, listing all qualifiers and class-name parameters. The usage summary explains how to invoke and exit from MONITOR, how to direct output, and any restrictions of which you should be aware. The Description Section explains how to use MONITOR. The Command Qualifiers Section describes each command qualifier. Qualifiers appear in alphabetical order. The Commands Section describes each MONITOR command. Commands appear in alphabetical order. The Examples Section contains examples of common operations that you perform with MONITOR. The Supplemental Information Section describes MONITOR recording file record formats. Associated Documents For additional information on the topics covered in this document, refer to the VAX/VMS DCL Dictionary and the VAX/VMS System Manager’s Reference Manual. Preface Conventions Used in This Document Convention Meaning A symbol with a one- to six-character abbreviation indicates that you press a key . on the terminal, for example, The phrase CTRL/x indicates that you must press the key labeled CTRL while you simultaneously press another key, for example, CTRL/C, CTRL/Y, CTRL/O. In examples, this control key sequence is shown as “x, for example, "C, *Y, "0, because that is how the system echoes control key sequences. $ SHOW TIME 05-MAY-1986 11:55:22 Command examples show in black letters all output lines or prompting characters that the system prints or displays. All user-entered commands are shown in red letters. $ TYPE MYFILE.DAT Vertical series of periods, or ellipsis, means either that not all the data that the system would display in response to the particular command is shown or that not all the data a user would enter is shown. file-spec.... Horizontal ellipsis indicates that additional parameters, values, or information can be entered. {logical-name] Square brackets indicate that the enclosed item is optional. (Square brackets are not, however, optional in the syntax of a directory name in a file specification or in the syntax of a substring specification in an assignment statement.) quotation marks apostrophes The term quotation marks is used to refer to double quotation marks (“). The term apostrophe (') is used to refer to a single quotation mark. New and Changed Features Version 4.4 of the Monitor Utility includes the following new functions: * A new CLUSTER class’ that permits live display and recording of significant clusterwide performance data for up to 16 member systems. Two screen formats are available. * A new CONVERT command that converts to Version 4.4 format recording files produced with previous MONITOR versions. * A new /NODE command qualifier' that allows users to request performance data for any or all classes for up to 16 VAXcluster member systems. * The I/O Request Queue Length item of the DISK class is now sampled every second regardless of the collection interval value. This change yields queue length statistics with a consistently low sampling error. 1 Requires that DECnet-VAX be installed. xi O MONITOR MONITOR The Monitor Utility (MVONITOR) is a system management tool that enables you to obtain information on operating system performance. FORMAT MONITOR [classname],...]] Command Qualifiers /BEGINNING=time Defaults (See Command Qualifiers Section.) /[NOJBY_NODE NOBY_NODE /[NOJDISPLAY[=file-spec] DISPLAY /FLUSH_INTERVAL=seconds /[NOJINPUT[=(file-spec,...)] /INTERVAL=seconds /NODE=(nodename,...) (See Command Qualifiers Section.) NOINPUT (See Command Qualifiers Section.) None. /[NOJRECORDJ=file-spec] /[NOJSUMMARY/-file-spec] NORECORD NOSUMMARY /[NOJCOMMENT="string" /ENDING=time /OUTPUT=file-spec /VIEWING_TIME=seconds NOCOMMENT (See Command Qualifiers Section.) None. (See Command Qualifiers Section.) Command Parameter classname],...] Specifies the class(es) of performance data to be monitored. To monitor all classes, specify the ALL_CLASSES parameter. When you specify several specific classes, separate the class-name parameters with commas or plus signs. Note that the ALL_CLASSES and CLUSTER class names are mutually exclusive. Note also that cluster monitoring functions require that DECnet— VAX be installed. MON-1 MONITOR ‘You must specify one or more of the following parameters: ALL_CLASSES Statistics for all classes CLUSTER Clusterwide performance statistics DECNET DECnet-VAX statistics DISK Disk 1/0 statistics DLOCK Distributed lock management statistics FCP File control primitive statistics FILE_SYSTEM_CACHE File system cache statistics 10 System 1/0 statistics LOCK Lock management statistics MODES Time spent in each of the processor modes PAGE Page management statistics POOL Statistics on space allocation in the nonpaged dynamic pool PROCESSES Statistics on all processes SCs System communications services statistics STATES Number of processes in each of the scheduler states SYSTEM Summary of statistics from other classes MONITOR class names accept a variety of qualifiers, as summarized in Table MON-1 in the Description Section. Complete descriptions of qualifiers for each class are provided in the Commands Section. usage summary Invoking To invoke MONITOR, issue the DCL command MONITOR. The utility then displays the following prompt: MONITOR> In response to the prompt, you can issue any utility command. Note that you can also initiate MONITOR requests from command level by issuing the DCL command MONITOR along with the desired qualifiers and parameters. However, in terms of system overhead, it is preferable to initiate requests in response to the MONITOR> prompt. Exiting Generally, each MONITOR request runs until the time specified or implied by the /ENDING qualifier. However, you can press CTRL/C or CTRL/Z to terminate a MONITOR request earlier. Pressing CTRL/C terminates the current request without exiting from the utility. You can then initiate a new request or issue any Monitor Utility command. Pressing CTRL/Z terminates the current request and exits from MONITOR. Directing Output Information collected by MONITOR is normally displayed as ASCII screen images. You can use the optional /DISPLAY qualifier to specify a disk file to contain the information. If you omit the file specification, output is directed to SYS$OUTPUT. See the Command Qualifiers Section for a discussion of the /DISPLAY qualifier. MON-2 MONITOR Description Privileges/Restrictions No privileges are needed to invoke MONITOR. monitor commands Syntax MONITOR> command MONITOR Commands CONVERT file-spec EXECUTE (@) file-spec EXIT HELP [command] INITIALIZE MONITOR classnamel,...] SET DEFAULT classname,...] SHOW DEFAULT e DESCRIPTION e e e The Monitor Utility permits you to monitor classes of systemwide performance data (such as system I/O statistics, page management statistics, and time spent in each of the processor modes) at specifiable intervals and to produce a variety of outputs. MONITOR collects system performance data by class and produces three forms of optional output: 1 A disk recording file in binary format 2 Statistical terminal displays 3 A disk file containing statistical summary information in ASCII format The utility initiates a single MONITOR request for the classes of performance data specified each time you issue a command of the form MONITOR [/qualifier(,...]] classname[,...] [/qualifier[,...]1] Regardless of the order in which you specify class-name parameters, MONITOR always executes requests in the following sequence: PROCESSES STATES MODES PAGE 10 FCP POOL LOCK DECNET FILE_SYSTEM_CACHE DISK DLOCK SCS SYSTEM Depending on the command qualifiers specified, MONITOR collects system performance data from the running system or plays back data recorded previously in a recording file. When you play back data, you can display it, summarize it, and even rerecord it to reduce the amount of data in the recording file. The Examples Section illustrates these operations in greater detail. MON-3 ) MONITOR Description For additional information on how to interpret the information the Monitor Utility provides, see the Guide to VAX/VMS Performance Management. 1 Class Types Each MONITOR class consists of various data items that, taken together, provide a statistical measure of a particular system performance category. The data items defined for individual classes are listed in the description of the MONITOR command in the Commands Section. There are two MONITOR class types, differentiated by the scope of the data items collected: ® System Classes, in which the data items provide, for the entire system, statistics on resource utilization (CLUSTER, DECNET, DLOCK, FCP, FILE_SYSTEM_CACHE, 10, LOCK, MODES, PAGE, POOL, STATES, SYSTEM). * Component Classes, in which the data items provide statistics on the contribution of individual components to the overall system or VAXcluster measure. These classes are DISK, PROCESSES, and SCS (system communication services). As an example of the distinction between MONITOR class types, the 10 class includes a data item to measure all direct I/O operations for the entire system. The DISK class, on the other hand, measures direct I/O operations for individual disks. Class-Name Qualifiers The class-name qualifiers control the type of display and summary output format generated for each class name specified. They have no effect on the recording of binary data. Each of these qualifiers applies only to the immediately preceding class name. Class-name qualifiers must not appear as part of the command verb. Table MON-1 summarizes class-name qualifiers and defaults. Table MON-1 MONITOR Class-Name Qualifiers Class Name Qualifiers Defaults ALL_CLASSES /ALL /AVERAGE /CURRENT See Commands Section /MAXIMUM /MINIMUM CLUSTER DECNET /ALL /AVERAGE /CURRENT /MAXIMUM /MINIMUM /CURRENT /ALL /AVERAGE /CURRENT JALL /MAXIMUM /MINIMUM DISK /ALL /AVERAGE /CURRENT /ITEM /MAXIMUM /MINIMUM /[NOJPERCENT /ALL /ITEM-OPERATION_RATE /NOPERCENT DLOCK /ALL /AVERAGE /CURRENT /MAXIMUM /MINIMUM JALL /ALL /AVERAGE /CURRENT JALL FCP /MAXIMUM /MINIMUM MON-4 MONITOR Description Table MON-1 (Cont.) MONITOR Class-Name Qualifiers Class Name Qualifiers Defaults FILE_SYSTEM_CACHE /ALL /AVERAGE /CURRENT JALL /MAXIMUM /MINIMUM 10 /ALL /AVERAGE /CURRENT /MAXIMUM /MINIMUM JALL LOCK /ALL /AVERAGE /CURRENT /MAXIMUM /MINIMUM JALL MODES /ALL /AVERAGE /[NOJCPU /CURRENT /CPU /CURRENT /NOPERCENT /MAXIMUM /MINIMUM /[NOJPERCENT PAGE /ALL /AVERAGE /CURRENT /MAXIMUM /MINIMUM JALL POOL /ALL /AVERAGE /CURRENT /ALL /MAXIMUM /MINIMUM PROCESSES /TOPBIO /TOPCPU /TOPDIO /TOPFAULT None SCSs /ALL /AVERAGE /CURRENT /ITEM /ALL /ITEM=KB_MAP /MAXIMUM /MINIMUM /[NOJPERCENT /NOPERCENT /ALL /AVERAGE /CURRENT /CURRENT /NOPERCENT STATES /MAXIMUM /MINIMUM /[NOJPERCENT SYSTEM /ALL /AVERAGE /CURRENT /CURRENT /MAXIMUM /MINIMUM The class-name qualifiers fall into three categories: 1 Statistics qualifiers (/ALL, /AVERAGE, /CURRENT, /MAXIMUM, and /MINIMUM) specify which statistics appear in display and summary output. These are conflicting qualifiers; specify no more than one of them with each class name in a MONITOR request. Note that statistics qualifiers cannot be used with the PROCESSES class name or for multifile summaries. 2 The data transformation qualifier (/[NOJPERCENT) controls whether data for the selected class name is expressed as percentages of a whole. This qualifier can be used only with the STATES, DISK, MODES, and SCS class names, and it is not allowed for multifile summaries. 3 Class-specific qualifiers (/CPU, /ITEM, /TOPBIO, /TOPCPU, /TOPDIO, and /TOPFAULT) control the output of a specific class. ® The /CPU qualifier is used specifically with the MODES class name to list CPU information for VAX-11/782 attached processor configurations (except for multifile summaries). * The /ITEM qualifier is used with the component statistics class names DISK and SCS to specify one or more data items for inclusion in display or summary output. * The /TOP qualifiers are used with the PROCESSES class name to produce bar-graphs showing the top processes, instead of the standard summary and display output. Top processes are the heaviest consumers of the resource being monitored. Up to eight processes can be shown in each display. Note that the /TOP qualifiers are mutually exclusive. Specify no more than one of them in a single request. MON-5 MONITOR Description Outputs The Monitor Utility can produce any combination of three forms of output for any single MONITOR request. The forms are display output, recording file output, and summary output. Output forms are specified with the /DISPLAY, /RECORD, and /SUMMARY qualifiers. ¢ /DISPLAY produces output in the form of ASCII screen images. Screen images are written at a frequency governed by the /VIEWING_TIME qualifier. * /RECORD produces a binary recording file containing data collected for requested classes; one record for each class is written per interval. * /SUMMARY produces an ASCII file containing summary statistics for all requested classes over the duration of the MONITOR request. If you specify /INPUT with any of these qualifiers, MONITOR collects performance data from one or more previously created recording files; otherwise, data is collected from counters and data structures on the running system. The MONITOR request begins and ends at times specified by the /BEGINNING and /ENDING qualifiers, respectively. 3.1 Display Output Display output consists of a series of terminal screen images. One screen image per requested class per requested viewing interval is produced. Any terminal supported by the VAX/VMS operating system with dimensions of at least 80 columns by 24 rows can be used. (You may have to issue the DCL command SET TERMINAL to set the proper dimensions.) Display output can also be routed to a file for subsequent printing. The amount of time between screen displays is determined by the /VIEWING_TIME value. Effective viewing time varies, however, depending on whether you are running MONITOR on your local system or on a remote node. For remote access, the time required to display the screen is included in the viewing time, while for local access, this time is not included. You will therefore probably want to use a larger viewing time than the 3-second default when running MONITOR on a remote system. The value appropriate for remote access depends on your terminal baud rate. For a 9600-baud terminal line, 6 seconds is a reasonable viewing time. For lower-speed lines, raise the viewing time appropriately. By pressing CTRL/W, you can temporarily override the /VIEWING_TIME value and generate a new display immediately following the current one. This feature is useful when the MONITOR display area has been overwritten by an operator message. You can also use CTRL/W in conjunction with a large /VIEWING_TIME value to generate display events on demand. MON-6 MONITOR Description 3.2 Display Data With the exception of the PROCESSES class, all displayable data items are rates or levels. Rates are shown in number of occurrences per second. A level is a value that indicates the size of the monitored data item. MONITOR can display any of four different statistics for each data item, as follows: ® Current rate or level * Average rate or level * Minimum rate or level * Maximum rate or level The last three statistics are measured since the beginning of the MONITOR request. The current statistic displays the most recently collected value for the rate or level. Any one or all four of the statistics can be requested. For certain classes, all the above statistics can be expressed as percentages. 3.3 Screen Formats With the exception of the PROCESSES, SYSTEM, and CLUSTER classes, there are two basic screen formats used for displaying MONITOR class data: the single-statistic screen and the multiple-statistic screen. The formats vary slightly depending on whether the class being displayed is a system or component class. Observe that there are three characteristics common to the screen formats: 1 The date and time appearing in the heading of each screen refer to the time at which the displayed data was originally collected. 2 The name of the node on which the data was originally collected also appears in the heading (except when playing back files that do not contain node name information, or when displaying CLUSTER class data). The node name is obtained from the SCSNODE system parameter or, if SCSNODE is null, from the SYS$NODE logical name established by DECnet. 3 The bottom line of the display is used for status information pertaining to the current MONITOR request. * If data collection is from a file of previously recorded monitor data, the word PLAYBACK appears at the left margin of the line. If, instead, the currently running system is being monitored, the word does not appear. ¢ If a summary file has been requested, the word SUMMARIZING e If creation of a recording file has been requested, the word RECORDING appears at the right margin of the line; otherwise, it appears in the middle of the line; otherwise, it does not appear. does not appear. MON-7 MONITOR Description 3.31 Single-Statistic Screen for System Classes This bar-graph style screen is used whenever one statistic (current, average, minimum, or maximum) is requested. Example MON-1 exhibits the maximum statistic for the STATES class. For other classes and statistics, the screen format remains the same, with different heading and data item descriptions. If the display of percentages is requested, the percent symbol (%) appears in the title and next to the numbers along the top of the graph. All examples in this screen format are rounded up or down to seven whole numbers (except percentages, which are truncated to three whole numbers). Example MON-1 Single-Statistic Screen VAX/VMS Monitor Utility PROCESS STATES on node SAMPLE 1 MAX | 16-MAY-1986 16:00:63 [ 10 20 30 P L Collided Page Wait Mutex & Misc Resource Wait Common Event Flag Wait Page Fault Wait Local Event Flag Wait Local Evt Flg (Outswapped) | 2 28 |ve |suvssarssnnarrsarsansarens 4 |enee 11 [ensrsnnssnn Hibernate (Outswapped) 2 1 Suspended (Outswapped) | 4 Compute (Outswapped) Current Process | ! 1 Free Page Wait te |eses 1 10 R T 3.3.2 + 1= 3 [ene Hibernate Suspended 40 T. Multiple-Statistic Screen for System Classes This tabular-style screen is used whenever all four statistics are requested with the /ALL class-name qualifier. Example MON-2 shows a multiple-statistic screen. The precision of the data items is seven whole and two decimal places. For each class, the screen format remains the same, with different heading and data item descriptions. If you request the display of percentages, as in the MODES class example in Example MON-3, the percent sign (%) appears in the title and the headings, and the figures consist of three whole and one decimal place. MON-8 MONITOR Desci Example MON-2 Sample Multiple-Statistic Scr VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS on node SAMPLE 3.33 3.16 22.50 0.83 Free List Fault Rate 26.33 15.68 Modified List Fault Ra Demand Zero Fault Rate te 4.68 12.00 Global Valid Fault Rat: o Wrt In Progress Fault System Fault Rate Free List Size Modified List Size Example MON-3 Rate 3.83 7.68 11.33 7.83 0.00 0.00 24.33 12.83 3366.00 3321.50 1.00 70.00 3. ow Page Write 1/0 Rate 45.00 1.68 Page Write Rate 18. 3281 45 1" 26 4 & Page Read 1/0 Rate 58 16.33 @ 18.00 8388 #8328 232888 AVE 38.33 rwwoo Fault Rate 16:13:38 CUR £8.00 ~3ro 16-MAY-1986 12. 11. 0. 24. 00 3366. 139. Sample Multiple-Statistic Screen (Data Expressed as Percentages) VAX/VMS Monitor Utility TIME IN PROCESSOR MODES (%) on node SAMPLE 16-MAY-1986 16:13:38 CUR% AVEY MIN% MAXY% Interrupt Stack 20.3 21.9 20.3 23.6 Kernel Mode 23.0 23.8 23.0 2.6 Executive Mode 3.0 3.5 3.0 24.6 Supervisor Mode 0.0 0.0 0.0 0.8 User Mode 51.3 46.9 42.8 61.3 Compatibility Mode 2.3 3.6 0.0 3.9 Idle Time 0.0 0.0 0.0 94.9 3.3.3 Component-Classes Screen For component classes, only one data item per component is displayed on each screen. The item is identified in the upper left of the screen. Components for which statistics are reported appear in the left column of the screens. If more than one item keyword is specified with the /ITEM qualifier, or if /ITEM=ALL is specified, a new screen appears for each item selected. For example, the command MONITOR DISK/ITEM=(OPERATION_RATE,QUEVE_LENGTH) would produce output of the format shown in Example MON-4. MON-9 MONITOR Description Example MON-4 Sample Component Statistics Screens VAX/VMS Monitor Utility DISK 1/0 STATISTICS on node SAMPLE 15-MAY-1086 20:08:42 1/0 Operation Rate DRA2: DRB1: DRC3: DRCA: DBA3: DBAG: DRAT: DUA4: DUAS: DUAT: SAMPLEPAGE ACCREG VMS_X20R SAMPLESECDO1 UMASTER MIDNITE RES26APR RES0BAUG VMSDOCLIB 0LD_QVsss CUR 0.00 0.00 1.99 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AVE MIN MAX 0.03 0.00 0.19 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.33 0.00 1.9 0.00 0.00 0.00 0.00 0.00 0.00 0.00 AVE 0.00 MIN 0.00 MAX 0.00 VAX/VMS Momitor Utility DISK 1/0 STATISTICS on node SAMPLE 16-MAY-1986 20:08:45 1/0 Request Queue Length DRA2: SAMPLEPAGE DRB1: DRC3: DRCA: DBA3: DBAG: DRAT: DUA4: DUAS: DUAT: 3.4 ACCREG VMS_X20R SAMPLESECDO1 UMASTER MIDNITE RES26APR RESOBAUG VMSDOCLIB 0OLD_QUss$ CUR 0.00 2.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1.43 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Recording File Output A recording file is a VAX RMS sequential disk file that is created when a MONITOR request includes the /RECORD qualifier. A record of binary performance data is written to this file once for each requested class per interval; the record contains a predefined set of data for each of the requested performance classes. The file is created when a MONITOR request is initiated and closed when the request terminates. The resulting file can be used as a source file by later requests to format and display the data on a terminal, to create a summary file, or to record a new recording file with different characteristics. All data pertaining to the class is recorded. Note that such is the case even if you are concurrently displaying only a single statistic or a single item of a component statistics class. MON-10 MONITOR Description 341 Disk Space for Recording Files When recording is active (or display output is being routed to a disk file), it is possible to consume large quantities of disk space in a short period of time. In particular, if disk quota is exceeded during execution of a MONITOR request, open files are closed and the request is terminated prematurely. To avoid this situation, carefully plan recording requests by estimating the amount of disk space required, using the following rules of thumb: * * Allow 70 bytes per class per interval or, for the PROCESSES class, 70 bytes per process per interval. For component classes, allow 4 bytes per interval for each line displayed when /ITEM=ALL is specified. When SYSTEM class data is recorded, the MODES, STATES, and PROCESSES classes are also recorded, even if not specifically requested. When CLUSTER class data is recorded, the MODES and DISK classes are also recorded. To estimate disk space requirements for CLUSTER recording files, multiply the amounts for these classes by the number of nodes being monitored. After estimating disk space requirements, check the amount of disk quota available, and set appropriate values for /INTERVAL and /ENDING. If necessary, refer to the Supplemental Information Section for details on the exact recording file record sizes. 342 3.5 Recording File Version Compatibility Before Version 4.4 MONITOR can read recording files generated by previous MONITOR versions, you must convert the files to Version 4.4 format. Use the CONVERT command described in the Commands Section of this document. Summary Output Summary output is an ASCII disk file consisting of one display screen image per requested class. The screen format for each class is based on the statistic requested. The only difference in format between a display screen and a summary screen image is that the word SUMMARY appears in the heading along with a beginning and ending time for the period covered by the summary. For all except the PROCESSES/TOP, SYSTEM, and CLUSTER summaries, the data contained in the summaries is identical to that shown on the final display screen (if display output was also requested). Since the summary file reflects the accumulation of data throughout the MONITOR request, the average, minimum, and maximum statistics are of particular interest. For the TOP summaries of the PROCESSES and SYSTEM classes, the data represents the top users for the entire duration of the MONITOR request, subject to the following restriction. To be eligible for inclusion in the list of top users, a process must be present and swapped in at the beginning and end of the MONITOR request. MON-11 MONITOR Description 3.6 Multifile Summaries Multifile summaries provide a convenient method of combining data from a number of recording files to compare AVERAGE performance statistics (excluding the PROCESSES and CLUSTER classes) for discrete time segments. You can use the /BEGINNING and /ENDING command qualifiers to delimit the desired time segment (see the Command Qualifier Section). To request a multifile summary, you use the /SUMMARY command qualifier and specify a list of recording files with the /INPUT qualifier. Note that since only AVERAGE statistics are collected, you should not specify class-name qualifiers. Note also that multifile summaries are “static"—that is, they do not provide continuously updated displays. Caution: Version 4.4 MONITOR file structure must be common to all recording files in the list. 3.6.1 Interpreting Multifile Summary Reports Multifile summary reports differ from regular (single-file) reports in both content (only AVERAGE statistics are collected) and format. MONITOR formats multifile report data as follows: * By file—This is the default format. For each class requested, the report displays one column of AVERAGE statistics per input file, along with beginning and ending times for each file. For files that contain data for multiple nodes, there is one column per node per file. * By node—To request this format, you specify the /BY_NODE command qualifier (along with the /SUMMARY and /INPUT qualifiers) when you create the summary file. The report combines data from all files for a given node into a single column that shows the average statistic for each data item. The contribution of the data from each file is weighted by the amount of time over which the data was collected (for rate items) or by the number of collections (for level items). For both formats, MONITOR provides “Row Sum,” “Row Average,” “Row Maximum,” and “Row Minimum” statistics. These represent simple arithmetic operations performed on all the averages in each row of the report. Note: Since in many cases multifile summary reports contain large amounts of data, and since they use a 132-character format, you will probably want to print them out. Moreover, a single page can display only five columns of data. Depending on the number of recording files, nodes, and classes specified, a report may extend over many pages. In that event, “Row” values appear on the final page. The following examples illustrate differences between single-file and multifile reports. First, consider two fragments of single-file reports generated on the same node from two different recording files, which cover, respectively, a two-hour and a ten-hour period. MON-12 MONITOR Description VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS on node BLUE From: 16-MAY-1986 09:00:00 SUMMARY CUR Page Fault Rate AVE To: 90.00 16-MAY-1086 11:00:00 MIN MAX VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS on node BLUE From: SUMMARY CUR Page Fault Rate To: AVE 16-MAY-1986 11:00:00 16-MAY-1986 21:00:00 MIN 6.00 MAX The corresponding “by node” multifile report fragment for the entire period shows that the average Page Fault Rate is weighted toward the figure that represents the larger elapsed time: VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS Multifile SUMMARY Node: From: To: BLUE (2) 16-MAY-1986 09:00 Row 20.00 20.0 16-MAY-1986 21:00 Page Fault Rate Sun Row Average 200 Row Minimum 20.00 Row Maximum 20.00 The “Row” statistics provided in multifile reports are not time-weighted; they are meaningful only when all input files that contribute to the report cover a common time period, as in Example MON-6 (“by node” report for nodes MOE, CURLEY, and LARRY) in Section 3.6.3. Thus, as the following example shows, “Row” statistics would not be useful in a “by file” report for node BLUE, because the contributing files covered different time periods: VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS Multifile SUMMARY BLUE BLUE From: 16-MAY-1986 09:00 16-MAY-1986 11:00 Row Row To: 16-MAY-1986 11:00 16-MAY-1086 21:00 Minimun Maximum Page Fault Rate 90.00 6.00 8.00 90.00 Those two examples illustrate how MONITOR formats multifile report data. Each column of averages is headed by the name of the node on which the data was collected, the requested beginning and ending times, and (for “by node” reports) a parenthesized number signifying the number of input files that contributed data to the column. If no explicit beginning and ending times are specified in the summary request, the times stored in the files are listed. Note that if a column in a “by node” report contains data from more than one input file, the earliest beginning time and the latest ending time among all the time stamps in those files are listed. You must therefore exercise caution in interpreting “by node” reports, because such reports may include in the listed time range a period during which no data was collected. MON-13 MONITOR Description 3.6.2 Using Multifile Summary Reports in Single-Node Environments In single-node environments, “by file” and “by node” reports have the following applications: * By file—You can use this type of report (which contains one column of average statistics per input file) when you want to compare data from several different time segments. The input files selected must already contain the time segments of interest when the multifile summary is run; the /BEGINNING and /ENDING qualifiers cannot be used for this purpose, because they can define only one time segment. If necessary, execute a preliminary MONITOR command to rerecord an input file and write data for the desired time segment to a new file. Example MON-5 at the end of this section shows a muiltifile report generated from three input files, each of which contained data for the period 8 A.M. to noon on a different day. Some of these files may have been created as a result of rerecording larger files to extract only the 8 AM. to noon time segment. Note that the average page fault rate for a morning during the week of 11 March was 28.4, and that the daily averages for that time period ranged from 21.74 to 39.53. * By node—If you specify the /BY_NODE qualifier, you can combine data from several input files to produce a single average statistic for each data item. If continuous recording is implemented, so that MONITOR data is recorded whenever the system is up, you can produce a report that shows averages for any time segment, regardless of how many input files are included (up to the maximum of 125). Use the /BEGINNING and /ENDING qualifiers to define the time segment of interest. Example MON-5 shows a typical “by file” multifile summary report for three input files on a single node. The summary was requested with the following command: MONITOR /INPUT=(11MAR,12MAR,13MAR)/NODISPLAY/SUNMARY PAGE Example MON-5 Sample Single-Node Muitifile Summary VAX/VMS Monitor Utility PAGE MANAGEMENT STATISTICS Multifile SUMMARY YELLOW : To: YELLOW YELLOW 11-MAY-1986 08:00 12-MAY-1986 08:00 13-MAY-1986 08:00 Row Row Row Row 11-MAY-1986 12:00 12-MAY-1986 13-MAY-1986 12:00 Sun Average Minimum Maximum Page Fault Rate Read Rate Read 1/0 Rate Page Write Rate 1! 39.53 23.98 21.74 86.2 28.4 21.74 39.53 14.12 3.30 2.21 10.6 6.5 2.21 14.12 1.73 0.52 0.43 2.7 0.9 0.43 1.73 11.21 1.54 1.11 13.7 4.8 0.0 111 0.00 11.21 20.0 6.8 4.73 10.28 4.9 4.53 4.21 11.58 Page Write 1/0 Rate 0.11 0.01 0.01 Free List Fault Rate 10.28 5.00 4.73 6.51 4.77 4.53 Modified List Fault Rate Demand Zero Fault Rate Global Valid Fault Rate Wrt In Progress Fault Rate System Fault Rate Free List Size Modified List Size MON-14 11.68 5.08 0.1 14.8 4.21 20.8 8.9 0.11 6.51 10.26 8.54 7.76 26.5 8.8 7.76 10.25 0.03 0.01 0.01 0.0 0.0 0.01 0.03 6820.80 7102.33 7271.856 21203.7 7067.9 6820.60 7271.85 0.04 267.31 0.09 210.41 0.06 216.91 0.2 684.6 0.0 228.2 0.04 210.41 0.09 267.31 MONITOR Description 3.6.3 Using Multifile Summary Reports in VAXcluster Environments In VAXcluster or other multinode environments (DECnet or any set of noncommunicating nodes), “by file” and “by node” reports have the following applications: * By file—You can use this type of report to compare several time segments as you would for the single-node case. By selecting the appropriate input files, you can include in the report data from any single node or set of nodes. ¢ By node—As in the single-node case, you can use this type of report to combine data from several input files for a given node in one column of averages. Note, however, that there is one column for each node. By specifying the /BEGINNING and /ENDING qualifiers to select a common time period, you can use the report to make node-to-node comparisons. In addition, you can examine the “Row” statistics to evalute overall cluster performance. For VAXcluster environments only, you can use the /NODE qualifier to select data for one or more nodes from the set of input files. Moreover, you can combine data for those nodes by specifying the /BY_NODE qualifier. In Example MON-6 below, consider the Idle Time item from the MODES class. The Row Sum figure (116.2) represents the average idle time for the entire cluster over the two-hour period covered by the report; the Row Average (38.7) is the average idle time per node over the same time period. The Row Minimum (0.02) and Row Maximum (92.63) figures identify the lowest and highest of the individual node averages. Example MON-6 shows a typical VAXcluster multifile summary. It was generated by the following command: MONITOR/INPUT=(MOE.DAT; %, CURLEY .DAT; * ,LARRY .DAT; *) MODES,PAGE - /SUMMARY/BY_NODE /NODISPLAY/BEGINNING="18:17"/ENDING="20:17" Sample command procedures for generating multifile summaries in a VAXcluster environment are provided in the Examples Section of this document. Error Messages The VAX/VMS System Messages and Recovery Procedures Reference Manual lists the messages generated by MONITOR and provides explanations and suggested user actions. MON-15 MONITOR Description Example MON-6 Sample VAXcluster Multifile Summary VAX/VMS Monitor Utility TIME IN PROCESSOR MODES MULTI-FILE SUMMARY MOE CURLEY (2) LARRY From: 16-NAY-1086 15-MAY-1086 18:17 Row Row Row Row To: 15-MAY-1986 20:17 15-MAY-1086 20:17 15-MAY-1086 20:17 Sun Average Minimun Maximum Interrupt Stack 18:17 15-MAY-1986 18:17 8.51 0.50 8.26 13.2 4.4 0.50 8.51 26.73 0.42 12.43 Executive Mode 9.48 0.96 1.81 38.5 12.8 0.42 26.73 Supervisor Mode 1.97 0.00 0.18 2.1 0.7 0.00 1.97 32.72 5.33 79.32 117.3 39.1 5.33 79.32 Kernel Mode User Mode 12.2 4.0 0.96 9.46 Compatibility Mode 0.00 0.07 0.00 Idle Time 23.61 92.63 0.02 118.2 15-MAY-1086 18:17 16-MAY-1086 18:17 15-MAY-1988 18:17 15-MAY-1086 20:17 16-MAY-1986 20:17 15-MAY-1988 20:17 Row Row Row Row Sum Average Minimum Maximum 0.0 0.0 38.7 0.00 0.02 0.07 92.63 VAX/VMS Monitor Utility | AVE | PAGE MANAGEMENT STATISTICS MULTI-FILE SUMMARY Node: From: To: MOE Page Fault Rate Page Read Rat Page Read I/0 Rate Page Write Rate Page Write I/0 Rate Free List Fault Rate Modified List Fault Rate CURLEY (2) 36.73 8.81 0.49 46.0 16.3 0.49 36.73 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 2.29 10.2 0.0 0.0 5.83 0.24 0.0 0.0 0.00 8.1 2.7 0.00 8.10 2.69 0.00 10.8 3.8 0.00 0.53 0.18 5.8 1.8 0.18 14.28 1.20 8.80 4.71 0.70 1.40 Demand Zero Fault Rate 12.96 Wrt In Progress 0.00 0.00 7686.30 8630.14 Global Valid Fault Rate Fault Rate Systen Fault Rate Free List Size Modified List Size MON-16 LARRY 4.92 87.69 1.68 324,07 0.00 0.00 0.24 0.00 9665.08 32.12 19.0 1.9 14.8 0.0 26881.5 443.8 6.3 0.6 3.4 4.9 0.0 8627.1 147.9 0.00 0.00 0.24 0.24 0.00 75686.30 32,12 14.28 1.20 8.60 5.83 12.98 8.10 0.00 4.92 9665.06 324.07 MONITOR Command Qualifiers COMMAND QUALIFIERS This section describes qualifiers for the MONITOR classname and SET DEFAULT commands. Note that these commands accept the same qualifiers. The qualifiers follow the standard rules of DCL grammar, as specified in the VAX/VMS DCL Concepts Manual. Thus, you can abbreviate any qualifier or keyword as long as the abbreviation is not ambiguous. The asterisk and the percent sign can be used as wildcard characters unless otherwise noted. MON-17 MONITOR /BEGINNING /BEGINNING Specifies the time that monitoring begins. FORMAT /BEGINNING=time time Specifies a combination of absolute and delta times. Observe the syntax rules for time values described in the VAX/VMS DCL Concepts Manual. DESCRIPTION If you are monitoring a running system and you omit the /BEGINNING qualifier, monitoring begins at the time you issue the MONITOR command. However, if you have also specified the /INPUT qualifier to play back data from an input recording file, /BEGINNING defaults to the beginning time recorded in the input file. If you specify /BEGINNING with a time, but are playing back a recording file, MONITOR selects whichever is later: the beginning time of the file or the beginning time you specify. If you are monitoring a remote node, the local node time is used to determine beginning time. If you specify a future time for a request to monitor a running system, MONITOR issues an informational message, and the process issuing the request hibernates until the specified time. This feature can be useful when you run MONITOR from a batch job. MON-18 MONITOR /BY_NODE /BY_NODE Specifies that performance class data in a multifile summary be displayed as a single column of AVERAGE statistics for each node. FORMAT /[INOJBY_NODE DESCRIPTION The /BY_NODE qualifier can be specified only in combination with the /SUMMARY qualifier. One column of AVERAGE statistics per node appears for each class requested. Multifile summaries include one column of AVERAGE statistics for each node requested in each input file. (For more information on multifile summaries, see Section 3.6.) MON-19 MONITOR /COMMENT /COMMENT Specifies an ASCII string to be stored in the output recording file. The string can contain up to 60 characters. FORMAT /COMMENT="string /NOCOMMENT “string” Is a comment string up to 60 characters in length. e — DESCRIPTION The /COMMENT qualifier is valid only when /RECORD is also specified. If you omit the qualifier or specify /NOCOMMENT, a string consisting of 60 blanks is stored in the recording file by default. When a recording file containing a comment is played back, the comment is included in the heading of the display or single-file summary. MON-20 MONITOR /DISPLAY /DISPLAY Specifies whether information collected by MONITOR is to be displayed as ASCII screen images, and optionally names the disk file to contain the output. FORMAT /DISPLAY [=file-spec] /NODISPLAY [=file-spec] Is the specification of the file to which output is directed. —_— DESCRIPTION If you omit the optional file specification, output is written to SYSSOUTPUT. By default, display output is produced. Note, however, that display output is never produced when multifile summary is requested. Sections 3.1 to 3.3 describe the display output formats. MON-21 MONITOR /ENDING /ENDING Specifies the time that monitoring ends. e e e FORMAT e e S /ENDING=time time Specifies a combination of absolute and delta times. Observe the syntax rules for time values described in the VAX/VMS DCL Concepts Manual. DESCRIPTION !f you are monitoring a running system and you omit the /ENDING qualifier, monitoring continues until you terminate the request with CTRL/C or CTRL/Z. If you have also specified the /INPUT qualifier, so that data is played back from an input recording file, /ENDING defaults to the ending time recorded in the input file. If you specify /ENDING with a time, but are playing back a recording file, MONITOR selects the earlier of the ending time of the file and the ending time you specify. For live requests, the local node’s time stamp is used to determine ending time. You can prematurely terminate a request, regardless of the value of the /ENDING qualifier, by pressing CTRL/C or CTRL/Z. To prematurely terminate a request running in a noninteractive process (that is, a batch job or a detached process or subprocess), issue the appropriate DCL command to terminate the process. MON-22 MONITOR /FLUSH_INTERVAL /FLUSH_INTERVAL Specifies the interval, in seconds, at which data collected by MONITOR (contents of MONITOR buffers) is written to disk. Values must be in the range of 1 through 9,999,999. The default interval is 300 seconds. FORMAT /FLUSH_INTERVAL=seconds seconds Is a value in the range of 1 through 9,999,999. DESCRIPTION If you are writing data to a shared recording file currently in use, you should specify a short interval to ensure that others accessing the file receive data that is as current as possible. The smaller the interval, the less data is lost if a system failure occurs while recording. MON-23 MONITOR /INPUT /INPUT Controls whether performance data is played back from one or more input files or collected from the running system. FORMAT /INPUT [=(file-spec,...)] /NOINPUT file-spec(....) Specifies one or more input files. If you specify more than one file, enclose the list in parentheses and separate the file specs with commas. Wildcard characters are allowed in the file specification. Caution: Version 44 MONITOR file structure must be common to all recording files in the list. DESCRIPTION With multiple input files, you must use the /SUMMARY qualifier. The maximum number of files MONITOR will accept for a multifile summary is 125. In a list of input files, any omitted segment of the file specification (name or type) is defaulted to the corresponding segment of the previous file specification. If you omit the file type, the default file type DAT is used. If you omit the file specification, MONITOR assigns the default filename MONITOR.DAT. The current device and directory defaults are applied. If you omit the qualifier, performance data is collected from the running system. MON-24 MONITOR /INTERVAL /INTERVAL Specifies the sampling interval between data collection events, recording events, and display events. FORMAT /INTERVAL=seconds seconds Is a value in the range of 1 through 9,999,999. DESCRIPTION To understand how to use the /INTERVAL qualifier, you must be aware of various events occurring periodically within a MONITOR request. These are the collection event, the recording event, and the display event. The /INTERVAL qualifier is the primary means of controlling the frequency of these events. When a collection event occurs, raw data for all requested classes is collected from the operating system or from a previously recorded file. When a recording event occurs, data for all requested classes is written to a recording file. When a display event occurs, a screen image is composed, for a single class, from the accumulated data collected for that class since the beginning of the MONITOR request. For live collection requests, a collection event is always followed immediately by a recording event (if requested). The frequency of collection-recording event pairs is controlled by the /INTERVAL qualifier, which specifies the number of seconds that must elapse between occurrences of the event pair. Display events occur asynchronously to collection-recording event pairs, at a frequency governed by the /VIEWING_TIME qualifier. For playback requests, a collection event occurs each time a new interval is encountered in the input file of previously recorded data. A recording event (if requested) does not necessarily follow immediately as in the case of live collection. Its frequency is still governed by the /INTERVAL qualifier; the specified /INTERVAL value is interpreted in terms of the /INTERVAL value specified when the input file was created. The new value must be an integral multiple of the original value. A recording event is then triggered every time an interval is encountered in the input file that is the appropriate multiple of the original interval. For playback requests, occurrences of display events (if requested) are indicated in exactly the same way as recording events— with the /INTERVAL qualifier—and immediately follow recording events (if both are specified). The actual length of time a displayed image remains on the screen is still specified with the /VIEWING_TIME qualifier, but, unlike the live collection case, this qualifier is not used to signal a display event. Table MON-2 summarizes which qualifiers cause the various MONITOR events. MON-25 MONITOR /INTERVAL Table MON-2 Relationship of MONITOR Command Qualifier to Event Event Live Collection Qualifier Playback Qualifier Collection /INTERVAL Original /INTERVAL value (from file) Recording /INTERVAL /INTERVAL Display /VIEWING _TIME /INTERVAL Note that, for live requests, the collection interval is defined as the number of seconds from the end of one collection event to the beginning of the next. A collection event includes collection for all requested classes on all nodes specified. (For multiple-node requests, a collection event must complete on all nodes before a new event is initiated.) So, the elapsed time from the beginning of one collection event to the beginning of the next is the interval value plus the time it takes to do the collection. For some requests, notably those including many classes or the PROCESSES, CLUSTER, or SYSTEM classes, collection time can be significant. For /INPUT requests, the interval value defaults to the value specified in the input recording file. The default for monitoring the running system is 3 seconds for all classes except ALL_CLASSES, CLUSTER, and SYSTEM, which have a default of 6 seconds. MON-26 MONITOR /NODE /NODE Specifies the nodes—up to 16 in a VAXcluster—for which data is to be collected. FORMAT /NODE=(nodename,...) nodename(,...) Specifies one or more node names. If you specify more than one name, separate the names with commas and enclose the list in parentheses. DESCRIPTION You can specify a maximum of 16 node names with the /NODE qualifier. If you specify multiple node names with multiple system classes, MONITOR displays one class at a time for each node. For example, the command MONITOR/NODE=(NODE_A NODE_B) STATES,MODES generates STATES data for NODE_A and NODE_B, and then MODES data. Note that if you specify the /ITEM class-name qualifier with the DISK or SCS classes, MONITOR displays one item at a time for each node in turn. MON-27 MONITOR /OUTPUT /OUTPUT Used with the CONVERT command, this qualifier specifies the name of the converted recording file. FORMAT /OUTPUT=file-spec file-spec Is the file specification of the output file. The default specification is MONITOR.DAT. File lists are not permitted. DESCRIPTION Recording files produced using earlier MONITOR versions must be converted to Version 4.4 format before they can be played back with Version 4.4 MONITOR. MON-28 MONITOR /RECORD /RECORD Specifies that a binary disk file be created containing all collected data for the request. FORMAT /RECORD/=file-spec] /NORECORD file-spec Is the name of the recording file. Note that recording is restricted to files on disks. No wildcard characters are allowed in the file specification. If you omit the file type, the default file type is DAT. If you omit the file specification, output is generated to a file named MONITOR.DAT, in the current default device and directory. If you specify an existing file but omit the version number, a new version of the file is created. e — DESCRIPTION The output consists of all data for the requested classes, regardless of the class-name qualifiers specified. Note that recording file output is not produced when a multifile summary is requested. By default, no recording file output is produced. For more information on MONITOR recording file output, see Section 3.4. Recording file record formats are described in the Supplemental Information Section. MON-29 MONITOR /SUMMARY /SUMMARY Specifies that an ASCII disk file be created containing summary statistics on all collected data for this request. FORMAT /SUMMARY [=file-spec] /NOSUMMARY file-spec Is the name of the summary file. If the optional file specification is omitted, it defaults to MONITOR.SUM. By default, no summary output is produced. S — DESCRIPTION The summary file, generated at the end of monitoring, contains one or more pages of output for each requested class. The format of each page is similar to that of display output, and is determined by the class-name qualifiers. The /ALL qualifier is applied to all class-names for which no other qualifier is specified. Summary output formats are described in Section 3.5; multifile summaries are described in Section 3.6. MON-30 MONITOR /VIEWING_TIME /VIEWING_TIME For /DISPLAY requests, this qualifier specifies the duration for each screen image display. FORMAT /VIEWING _TIME=seconds seconds Is a value in the range of 1 through 9,999,999. DESCRIPTION If you are monitoring the running system, /VIEWING_TIME defaults to the /INTERVAL value. If you specify /INPUT and you are monitoring a recording file, /VIEWING _TIME defaults to 3 seconds. Effective viewing time varies, however, depending on whether you are running MONITOR on your local system or on a remote node. For remote access, the time required to display the screen is included in the viewing time, while for local access, this time is not included. You will therefore probably want to use a larger viewing time than the 3-second default when running MONITOR on a remote system. The value appropriate for remote access depends on your terminal baud rate. For a 9600-baud terminal line, 6 seconds is a reasonable viewing time. Note also that the time between screenfuls of data for the PROCESSES display is controlled by this qualifier. MON-31 MONITOR Commands COMMANDS This section describes MONITOR commands. For commands that specify class-name parameters (other than ALL_CLASSES), a sample display or summary of each class is provided, along with a brief description of the items in the class. The commands follow the standard rules of DCL grammar, as specified in the VAX/VMS DCL Concepts Manual. Thus, you can abbreviate any command as long as the abbreviation is not ambiguous. Note, however, that use of the RETURN key is not permitted within command lines entered in response to the MONITOR> prompt. MONITOR recognizes the exclamation point (!) as a comment character. Thus, full- or partial-line comments are acceptable in command files specified as input to MONITOR. MON-32 MONITOR CONVERT CONVERT Converts a pre-Version 4.4 MONITOR recording file to Version 4.4 format. FORMAT CONVERT file-spec command parameter Specifies the file to be converted. The default file specification is command qualifier file-spec MONITOR.DAT /OUTPUT Is the file specification of the converted file. The default specification is MONITOR.DAT. DESCRIPTION You must convert pre-Version 4.4 recording files to Version 4.4 format before attemping to play them back with Version 4.4 MONITOR. EXAMPLE MONITOR> CONVERT 24MAY_MONITOR.DAT/QUTPUT=24MAY_NEWMON .DAT The command in this example converts the file 24MAY_MONITOR.DAT to Version 4.4 format and names the output file 24MAY_NEWMON.DAT. MON-33 MONITOR EXECUTE (@) EXECUTE (@) Executes a series of Monitor Utility commands contained in a file. FORMAT EXECUTE (@) command file-spec Specifies a command file to be executed by the EXECUTE (@) command. parameter file-spec command qualifiers None. DESCRIPTION With the EXECUTE command, you can direct MONITOR to obtain command input from a specified file rather than from the terminal. The file may contain any valid Monitor Utility command except an EXECUTE (@) command. Commands in the file are executed sequentially. If you omit the optional file specification, the default is MONITOR.MON. After the file has executed, subsequent commands are obtained from the terminal. EXAMPLE MONITOR> EXECUTE INGMEM.MON MONITOR> MONITOR /RECORD Contents of the file INQMEM.MON are ! This file ! INTERVAL= ts defaults for a memory management inquiry using GE, 10, and PROCESSES/TOPFAULT ' SET DEFAULT /INTERVAL=6 PAGE, 10, PROCESSES/TOPFAULT In this example, appropriate default values for a memory management investigation are established in the file INQMEM.MON, and the file is executed with the EXECUTE command. Then a subsequent MONITOR command uses those defaults, adding the /RECORD qualifier, to display and record the selected classes with a 5-second interval. Note that the defaults established when the file INQMEM.MON is executed remain in effect until changed explicitly or until the utility is exited. MON-34 MONITOR EXIT EXIT Terminates MONITOR, returning control to command level. FORMAT EXIT command None. parameters command None. qualifiers MON-35 MONITOR HELP HELP The HELP command displays HELP information on MONITOR. FORMAT HELP [command] command command parameters Specifies the name of a MONITOR command for which HELP is desired. command qualifiers None. EXAMPLE MONITOR> HELP INITIALIZE INITIALIZE Re-establish initial default values for MONITOR qualifiers and class-name paraneters. Topic? The command in this example requests help information on the INITIALIZE command. MON-36 MONITOR INITIALIZE INITIALIZE The INITIALIZE command reestablishes initial default settings for parameters and qualifiers previously altered by the SET DEFAULT command. FORMAT INITIALIZE command None. parameters command None. qualifiers MON-37 MONITOR MONITOR ALL_CLASSES MONITOR ALL_CLASSES The MONITOR ALL _CLASSES command initiates monitoring of statistics for all classes except the CLUSTER class. _—ms FORMAT MONITOR ALL_CLASSES command /qualifierl,...] qualifiers class-name qualifiers One or more qualifiers as described in the Command Qualifiers Section. /ALL Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION If you do not specify any qualifiers with the ALL_CLASSES parameter, normal default output is produced for each class. (See Table MON-1.) The qualifiers have no effect on display of the PROCESSES class. Note that the default interval is six seconds. The MONITOR ALL _CLASSES command is particularly useful for playback of recording files, since it eliminates the need to specify the particular classes of performance data the recording file contains. To override any of the degult qualifiers, specify the class name with the qualifier after specifying ALL_CLASSES. MON-38 MONITOR MONITOR ALL_CLASSES EXAMPLE MONITOR> MONITOR/INPUT=SYS$MANAGER:LOADBAL.DAT ALL_CLASSES,PROCESSES/TOPCPU This command initiates playback of the recording file SYS$MANAGER:LOADBAL.DAT. All data contained in the file will be displayed. MON-39 MONITOR MONITOR CLUSTER MONITOR CLUSTER Initiates monitoring of the CLUSTER statistics class, which shows clusterwide CPU, memory, disk, and locking activity. ————————————— _—— FORMAT MONITOR CLUSTER command qualifiers /qualifier,...] class-name /ALL qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. L DESCRIPTION For the CLUSTER class, MONITOR collects data items for up to 16 nodes in a VAXcluster. Because this class combines the most significant clusterwide performance statistics in a single display, it is particularly useful to cluster managers and other users seeking a general overview of cluster activity. Note that MONITOR does not recognize nodes that enter the cluster while a request is active. MONITOR will therefore not collect data for such nodes. Note also that the ALL _CLASSES and CLUSTER classes are mutually exclusive. MON-40 MONITOR MONITOR CLUSTER The CLUSTER class includes the following data items: ¢ CPU Busy—Percentage of CPU in use; includes activity in all processor modes (except Idle Time) for each node * Percent Memory In Use—Memory in use on each node; calculated by dividing the Free List Size by total available memory and subtracting the result from 100% ¢ 1/0 Operation Rate—Total rate of disk I/O operations on each disk by all nodes currently active in the request. Note that for MSCP-served disks, the operation rate includes all operations initiated on requested remote nodes, as well as all operations issued on the serving node to support remote operations. ¢ Total ENQ/DEQ Rate—Sum of all local, incoming, and outgoing ENQ’s, DEQ's, and conversions Two display formats are provided, depending on the class-name qualifier specified: ¢ A tabular style format for the /ALL qualifier ® A bar-graph style format for the /AVERAGE, /CURRENT, /MAXIMUM, and /MINIMUM qualifiers Note to Cluster Managers on MONITOR_SERVER Process When users issue the MONITOR CLUSTER command, MONITOR activates the image SYS$SYSTEM:VPM.EXE, which creates a process called MONITOR_SERVER on each active cluster node. (If users specify the /NODE qualifier with the MONITOR CLUSTER command, or with any command of the form MONITOR classname, MONITOR creates the process only on the specified nodes.) The purpose of the server process is to gather data from remote nodes for live display or recording on the local node. To ensure accurate and timely data collection, the process is started at priority 15. Since server processes consume minimal resources, they have no significant effect on system performance. By default, MONITOR_SERVER processes are started in the system DECNET account, which is created when the NETCONFIG.COM command procedure executes at bootstrap time. If for some reason this account is not present on your system, you must either create it by executing NETCONFIG.COM, or specify another account in which the server processes can be started. If you want to start the processes in another account, use the following sequence of commands to define VPM as known object 51 in the DECnet database and associate the object with the desired account. MON-41 MONITOR MONITOR CLUSTER $ SET PROCESS/PRIVILEGE=SYSPRV $ RUN SYS$SYSTEM:NCP NCP> DEFINE OBJECT NUMBER 61 - _ NAME VPN - FILE SYS$SYSTEM:VPM.EXE ~ PROXY NONE - _ ACCOUNT accountname - _ USERNAME username _ PASSWORD password NCP> SET OBJECT NUMBER 61 - _ NAME VPH _ FILE SYS$SYSTEM:VPM.EXE - PROXY NONE ACCOUNT accountname — USERNAME username _ PASSWORD password NCP> EXIT $ SET PROCESS/PRIVILEGE=NOSYSPRV For each server process, MONITOR creates on the local node a log file to which information on server connection activity, including error messages, is written. Note that error messages are written to the file only when errors occur, and that a single version is maintained for the life of the system. The default file specification has the form SYS$COMMON:[SYSMGR]VPM$nodename.LOG. The nodename portion of the specification identifies the node on which the MONITOR_SERVER process has been started. If you want to change the default specification, you can redefine the executive-mode logical name VPM$LOG _FILE in the system logical name table on the appropriate nodes. For example, if you wanted to write server error logging data to the file WRKD:[MONSERVER]VPM_ERRORS.LOG, you would define VPM$LOG _FILE as follows: $ DEFINE/SYSTEM/EXECUTIVE_MODE VPMS$LOG_FILE _$ WRKD: [MONSERVER] VPM_ERRORS .LOG To direct to a single file data for all MONITOR_SERVER processes on the cluster, you could assign the logical name the same value on each member system. Note that because the log files are created as shared sequential files, multiple server processes can access them simultaneously. If you routinely monitor your cluster, you can reduce server startup time significantly by creating MONITOR _SERVER processes on each member node at bootstrap time and maintaining the processes for the life of the system. To do so, add the following lines to the appropriate site-independent startup command files: $ DEFINE/SYSTEM/EXECUTIVE_MODE VPM$SERVER_LIVE TRUE $ RUN/DETACH SYS$SYSTEM:VPM.EXE You can issue these commands interactively at any time. Note, however, that you require the following privileges: ALTPRI, NETMBX, PSWAPM, SYSNAM, SYSPRV, and TMPMBX. MON-42 MONITOR MONITOR CLUSTER EXAMPLES vonrrom> vonrro cuusteR/ALL VAX/VNS Monitor Utility CLUSTER STATISTICS on node CURLEY LARRY AVE MIN 100.00 100.00 99.83 100.00 8.52 8.50 8.52 100.00 100.00 g CUR CPU Busy 88& 5-MAY-1986 12:25:13 VAX/VNS Monitor Utility CLUSTER STATISTICS on node CURLEY 88.00 MOE 78.00 LARRY 72.00 CURLEY AVE 88.00 78.00 72.50 MIN 88.00 77.00 72.00 s38 CUR AMemory In Use 888 & 6-MAY-1086 12:25:10 VAX/VMS Monitor Utility CLUSTER STATISTICS (DECEIT) TSDPERF 0.32 0.53 $1118DUA3: (DECEIT) DUMPDISK 0.00 : (DECEIT) BPMDISK $1118DUA2: $1118DUAG: LARRY$DRA3 : MOE$DMA MOE$DRA! LARRY$DRA4 LARRY$DBBS3 : (DECEIT) PAGESWAPDISK QUALQUEST UVMEQAR USEROL TIMEDEV REGLIB 1.44 0.00 0.00 0.00 0.00 0.00 0.21 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 (DECEIT) ORLEAN 0.00 (DECEIT) MPISDATA 0.00 (DECEIT) QMISDATABASE 0.96 0.00 0.00 0.00 MI 8383888888258 82 & AVE 6.563 1.07 SQMCLUSTERVA QUALD oo00000000mmnd CUR 0.48 1.03 (DECEIT) (DECEIT) 8888888888588% 1/0 Operation Rate $1118DUA7: $1118DUAB: = on node CURLEY 5-MAY-1086 12:26:26 eooo0000000000 0 VAX/VMS Monmitor Utility CLUSTER STATISTICS on node CURLEY 5-MAY-1088 12:25:66 Tot ENQ/DEQ Rate MOE LARRY CURLEY CUR 7.90 20.48 1.93 AVE 14.92 14.64 13.20 MIN MAX 0.00 0.00 43.12 0.00 67.30 46.92 This example shows the tabular style format for the CLUSTER display. MON—43 MONITOR MONITOR CLUSTER B MONITOR> MONITOR CLUSTER 11-MAY-1086 14:32:14 VAX/VNS Monitor Utility Statistic: CURRENT CLUSTER STATISTICS MEMORY | cPU | 100|#Menory In Use CPU Busy CURLEY LARRY MOE 68 |sonnsurssrsan 22 [ener 0 26 60 + S B2 |sersrsarae 41 [serannss 04 [erwennenens O1 [eessrnenens 12 |*s DISK 1/0 Operation Rate O 76 LOCK | | 100|Tot ENQ/DEQ Rate 0 125 S Hmmmmbm |CURLEY 103 |4wes ILARRY 36 Is IMOE 260 376 500 e benn et 82 |*#s This example shows the bar-graph style format for the CLUSTER display. MON-44 MONITOR MONITOR DECNET MONITOR DECNET The MONITOR DECNET command initiates monitoring of the DECNET class, which includes information on DECnet-VAX network activity. FORMAT MONITOR DECNET command qualifiers /qualifierl,...] class-name qualifiers /ALL One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. e DESCRIPTION The DECNET class consists of the following data items: ® Arriving Local Packet Rate—Rate at which local packets are being received ® Departing Local Packet Rate—Rate at which local packets are being sent * Arriving Transit Packet Rate—Rate at which transit packets are arriving ® Transit Congestion Loss Rate—Rate of transit congestion loss ® Receiver Buff Failure Rate—Rate of receiver buffer failures ¢ LRPs Available— Number of large request packets not in use MON-45 MONITOR MONITOR DECNET EXAMPLE MONITOR> MONITOR DECNET VAX/VMS Monitor Utility DECNET STATISTICS on node SAMPLE 16-MAY-1986 22:22:44 Arriving Local Packet Rate CUR 9.54 AVE 5.08 MIN 0.00 MAX 11.26 Arriving Trans Packet Rate 0.00 0.00 0.00 0.00 0.00 0.00 Departng Local Packet Rate Trans Congestion Loss Rate Receiver Buff Failure Rate LRPs Available 9.22 0.00 0.00 13.00 4.68 0.00 11.50 0.00 0.00 9.00 10.92 0.00 0.00 16.00 This example shows that arriving and departing network packet rates (including control packets) are roughly equivalent, and that network activity is currently at a level higher than the average since monitoring began, but not at its highest point. Note also that the count of LRPs is displayed; LRPs are used by various components of VAX /VMS, but primarily by DECnet— VAX. A sufficient number of these large request packets must be available to ensure that at peak periods of network use they are not exhausted. If they become depleted, network performance may degrade. The number of packets preallocated at boot time is determined by the SYSGEN parameter LRPCNT. MON-46 MONITOR MONITOR DISK MONITOR DISK The MONITOR DISK command initiates monitoring of the DISK statistics class. FORMAT MONITOR DISK command /qualifierl,...] qualifiers class-name qualifiers One or more qualifiers as described in the Command Qualifiers Section. /ALL Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /ITEM=(keyword],...]) Selects one or more data items for inclusion in display and summary output. If you specify two or more keywords, enclose them in parentheses and separate them with commas. When the /ITEM qualifier is omitted, the default is /ITEM=OPERATION _RATE. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. /[NOJPERCENT Controls whether statistics are expressed as percent values in display and summary output. The /PERCENT qualifier is applicable only to the DISK, MODES, SCS, and STATES classes. By default, statistics are not expressed as percent values. MON-47 MONITOR MONITOR DISK /ITEM Qualifier Keywords ALL Specifies that statistics on all data items collected for the disks are displayed on successive screens. OPERATION_RATE Specifies that I/O operation rate statistics are displayed for each disk. QUEUE_LENGTH Specifies that the number of 1/0 request packets being serviced (current or waiting) is displayed for each disk. e DESCRIPTION The DISK class is a component class. Data items for this class are collected for each mounted disk device in a single-node or VAXcluster system. The DISK class consists of the following data items: ® 1/0 Operation Rate—Rate at which I/O operations occur on each disk. By comparing operation rates for all disks in the system, you can discern which disks are busy and which are idle. However, since this statistic does not provide information on the time required for individual operations, you must use discretion in interpreting it. * 1/0 Request Queue Length—Number of outstanding I/O request packets. Includes the request currently being serviced and those awaiting service. Note that this item is always sampled at a 1-second interval, regardless of the value specified with the /INTERVAL command qualifier. In the following example, typical of a VAXcluster environment, note that each disk is identified by three elements: 1 The disk name ending in a colon. 2 The name of the VAXcluster node through which the disk is accessed. This field appears only in the multiple-statistic display; it is not included in single-statistic displays or multifile summaries. 3 The volume label. EXAMPLE MONITOR> MONITOR DISK/ITEM=QUEUE_LENGTH VAX/VMS Monitor Utility DISK 1/0 STATISTICS on node SAMPLE 16-MAY-1986 14:19:56 1/0 Request Queue Length CUR AVE MIN MAX SAMPLE$DBAO: SAMPLEO9APR SAMPLE$DRA2: SAMPLEPAGE SAMPLE$DRB1 : ACCREG $18DRAS: (MOE) ~ MOE$$PAGE $18DBA3: (CURLEY) UMASTER $18DBAG: (CURLEY) MIDNITE $28DRAT: (LARRY) RES26APR 0.00 2.00 0.00 0.00 0.00 0.00 0.00 0.00 1.43 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 4.00 0.00 0.00 0.00 0.00 0.00 $265$DUA4: 0.00 0.00 0.00 0.00 $2$DRB6: (LARRY) $2558DUAG: (SHEMP) VMSDOCLIB (SHEMP) CLUSTERDUMP1 RESOBAUG 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 This example, typical of a VAXcluster environment, shows, for each disk, the number of /O packets awaiting service or in service. Note that the MON-48 MONITOR N\ MONITOR DISK device SAMPLE$SDRA? is the only device with a nonzero queue length. Since MONITOR samples queue lengths every second, regardless of the collection interval value, the accuracy of the data does not depend on the collection —~ interval. MON-49 MONITOR MONITOR DLOCK MONITOR DLOCK The MONITOR DLOCK command initiates monitoring of the DLOCK (distributed lock management) statistics class. FORMAT MONITOR DLOCK command qualifiers /qualifier],...] One or more qualifiers as described in the Command Qualifiers Section. class-name /ALL qualifiers Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The DLOCK class is useful for monitoring the lock management subsystem in a VAXcluster environment. The class consists of the following data items: * New ENQ Rate (Local)—Rate of new lock (ENQ) requests that originate and are performed on this system MON-50 * New ENQ Rate (Incoming)—Rate of new lock requests that originate on other systems and are performed on this system * New ENQ Rate (Outgoing)—Rate of new lock requests that originate on this system and are performed on another system * Converted ENQ Rate (Local)—Rate of lock (ENQ) conversion requests that originate and are performed on this system MONITOR MONITOR DLOCK ¢ Converted ENQ Rate (Incoming)—Rate of lock conversion requests that originate on other systems and are performed on this system ¢ Converted ENQ Rate (Outgoing)—Rate of lock conversion requests that originate on this system and are performed on another system * DEQ Rate (Local)—Rate of unlock (DEQ) requests that originate and are performed on this system ¢ DEQ Rate (Incoming)—Rate of unlock requests that originate on other systems and are performed on this system * DEQ Rate (Outgoing)—Rate of unlock requests that originate on this system and are performed on another system * Blocking AST Rate (Local)—Rate of lock manager blocking ASTs that originate and are performed on this system * Blocking AST Rate (Incoming)—Rate of lock manager blocking ASTs that originate on other systems and are performed on this system e Blocking AST Rate (Outgoing)—Rate of lock manager blocking ASTs that originate on this system and are performed on another system * Directory Function Rate (Incoming)—Rate of requests for locks being managed by this node ¢ Directory Function Rate (Outgoing)—Rate of requests for locks being managed by other nodes ¢ Deadlock Message Rate—Rate of incoming and outgoing messages required for deadlock detection EXAMPLE MONITOR> MONITOR DLOCK VAX/VNS Monitor Utility DISTRIBUTED LOCK MANAGEMENT STATISTICS on node SAMPLE 16-MAY-1086 New ENQ Rate (Local) (Incoming) AVE MIN MAX 16.84 1.67 11.59 2.62 1.54 0.11 26.88 26.06 4.48 6.71 0.00 70.19 (Outgoing) Converted ENQ Rate (Local) 0.06 23.67 (Outgoing) 0.00 (Incoming) DEQ Rate 11:02:20 CUR 0.63 9.13 1.43 (Local) 15.86 11.58 (Outgoing) Blocking AST Rate (Local) 0.0 0.00 0.63 0.00 (Outgoing) Dir Functn Rate (Incoming) (Outgoing) 0.00 8.00 1.00 (Incoming) (Incoming) Deadlock Message Rate 1.68 0.00 0.00 0.00 0.99 0.00 26.68 0.00 0.00 5.99 0.01 0.00 0.00 0.00 0.00 16.90 1.64 2.59 0.00 7.33 0.77 5.99 41.22 0.00 4.68 0.00 0.00 24.85 0.00 0.00 11.00 2.66 0.00 This example shows that most of the current lock management activity occurs locally, but that, at some point during the monitoring period, there was a significant amount of incoming activity. MON-51 MONITOR MONITOR FCP MONITOR FCP The MONITOR FCP command initiates monitoring of the File Control Primitive statistics class, which includes information on all Files—11 Ancillary Control Processes (ACPs) and Extended QIO Programs (XQPs) on the local node. FORMAT MONITOR FCP command /qualifier],...] class-name /ALL qualifiers qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. e DESCRIPTION MON-52 The ECP class consists of the following data items, all of which are displayed as occurrences per second: * FCP Call Rate—Rate of QIO requests received by the file system. * Allocation Rate—Rate of calls that caused allocation of disk space. * Create Rate—Rate at which new files were created. * Disk Read Rate—Rate of read 1/0 operations from disk by the file system. * Disk Write Rate—Rate of write I/O operations to disk by the file system. MONITOR MONITOR FCP ® Volume Lock Wait Rate—Rate of entry into a wait state due to contention for a volume synchronization lock. Volume synchronization locks are taken out by the XQP during file creation, deletion, extension, and truncation operations. ® CPU Tick Rate—Rate at which CPU time was used by the file system (in 10-millisecond ticks). ® File System Page Fault Rate—Rate at which page faults occurred in the file system. ¢ Window Turn Rate—Rate of file-map window misses. ¢ File Lookup Rate—Rate of filename look-up operations in file directories. * File Open Rate—Rate at which files were opened. ¢ Erase Rate—Rate of erase operations issued by the file system. EXAMPLE MONITOR> MONITOR /INTERVAL=10 FCP VAX/VMS Monitor Utility FILE PRIMITIVE STATISTICS on node SAMPLE 16-MAY-1986 16:13:38 CUR FCP Call Rate Allocation Rate Create Rate Disk Read Rate Disk Write Rate Volume Lock Wait Rate CPU Tick Rate File Sys Page Fault Rate Window Turn File Lookup Rate File Open Rate Erase Rate 4.62 0.99 2.31 1.08 3.30 4.62 3.63 0.00 1.98 0.33 2.00 0.00 AVE 3.80 0.24 0.57 2.48 2.30 3.08 3.88 0.00 0.99 1.40 3.54 0.00 MIN 0.33 0.00 0.00 0.33 0.33 0.00 0.33 0.00 0.00 0.00 2.00 0.00 MAX 7.61 0.99 2.31 6.95 5.62 6.95 10.26 0.00 1.98 4.63 5.10 0.00 This example shows that the rate of file opens during the last 10-second collection interval was 2.0 (for a total of 20). The average rate since the MONITOR command was issued is 3.54; the highest rate achieved during any 10-second interval is 5.10, and the lowest rate of 2.0 occurred during the last interval. MON-53 MONITOR MONITOR FILE_SYSTEM_CACHE MONITOR FILE_SYSTEM_CACHE The MONITOR FILE_SYSTEM_CACHE command initiates monitoring of the FILE_SYSTEM_CACHE statistics class. FORMAT MONITOR FILE_SYSTEM_CACHE command /qualifier],...] class-name JALL qualifiers qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The FILE_SYSTEM..CACHE class includes the following data items: * Directory FCB Hit%-—Percentage of directory file control block hits on the Directory Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. * Directory FCB Attempt Rate—Rate at which attempts were made to find directory file control blocks in the Directory Cache. * Directory Data Hit%—Percentage of directory data hits on the Directory Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. * MON-54 Directory Data Attempt Rate—Rate at which attempts were made to find directory data in the Directory Cache. MONITOR MONITOR FILE_SYSTEM_CACHE ¢ File Header Hit%—Percentage of file header hits on the File Header Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. ¢ File Header Attempt Rate—Rate at which attempts were made to find file headers in the File Header Cache. File ID Hit%—Percentage of file identifier hits on the File ID Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. ¢ File ID Cache Attempt Rate—Rate at which attempts were made to find * Extent Cache Hit%—Percentage of appropriate size extent hits on the Extent Cache. The percentage value shown is the ratio of hits to the sum file identifiers in the File ID Cache. of hits plus misses. * Extent Cache Attempt Rate—Rate at which attempts were made to find appropriate size extents in the Extent Cache. * Quota Cache Hit%—Percentage of quota entry hits on the Quota Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. * Quota Cache Attempt Rate—Rate at which attempts were made to find * Bitmap Cache Hit%—Percentage of entry hits on the Bitmap Cache. The percentage value shown is the ratio of hits to the sum of hits plus misses. * Bitmap Cache Attempt Rate—Rate at which attempts were made to find entries in the Quota Cache. entries in the Bitmap Cache. Note that all items shown in the FILE_SYSTEM_CACHE display except Dir FCB apply only to XQPs. The Dir FCB item applies to both XQPs and the ODS-1 ACP. S S —— EXAMPLE MONITOR> MONITOR FILE_SYSTEM_CACHE VAX/VMS Monmitor Utility FILE SYSTEM CACHING STATISTICS on node SAMPLE 15-MAY-1986 13:08:63 CUR Dir FCB (Bit %) Dir Data (Attempt (Hit %) File Hdr (Bt %) File ID (Bit %) 100.00 AVE 100.00 MIN 0.00 0.49 100.00 0.00 0.00 (Attempt Rate) 66.00 1.00 80.00 0.41 0.00 0.00 Rate) 0.00 0.00 0.00 Rate) 0.00 0.24 0.00 (Attenpt (Attempt Extent (Hit %) Quota (Hit %) Bitmap (Hit %) (Attempt Rate) Rate) (Attempt Rate) (Attempt Rate) 1.68 100.00 4.68 0.00 1.24 0.00 0.00 100.00 0.00 0.00 0.00 0.00 100.00 0.18 0.00 0.00 0.00 0.00 MAX 100.00 1.68 100.00 4.68 100.00 1.00 0.00 0.00 0.00 100.00 0.00 0.00 0.00 0.00 100.00 0.86 0.00 0.00 1.00 The cache hits and misses reflect the effectiveness of file system caching. Generally, the size of the cache affects the hit rate. The Attempt Rate is the MON-55 MONITOR MONITOR FILE_SYSTEM_CACHE sum of hits plus misses; the Hit% is the percentage of attempts that were successful. Unlike other Monitor data items, the averages for the hit percentages are not calculated based on previous hit percentages. Instead, these values are calculated based on the total number of hits and the total number of attempts on a cache since the beginning of the Monitor request. The reason for this difference is to provide more accurate average values for the hit percentage items. The Directory FCB Cache is checked whenever a directory lookup is performed. Directory lookups may be performed on file open, creation, deletion, extension, or truncation. If the file control block associated with the directory is found in the cache, a hit is recorded. Otherwise a miss is recorded. Both hits and misses are counted as attempts. The Directory Data Cache is checked whenever a file lookup is performed. Directory lookups may be performed on file open, creation, deletion, extension, or truncation. If an entry for the file being accessed is found in the directory data cache, a hit is recorded. Otherwise a miss is recorded. Both hits and misses are counted as attempts. The File Header Cache is checked on file open, close, creation, deletion, extension, or truncation. If the file header for the file being accessed is found in the File Header Cache, a hit is recorded. Otherwise a miss is recorded. Both hits and misses are counted as attempts. The File Identification Cache is a list of file identifiers that are removed on file creation and returned on file deletion. The File Id hits indicate file numbers successfully removed or returned to the File Id Cache. Otherwise a miss is recorded. Both hits and misses are counted as attempts. The Extent Cache is checked on file creation, deletion, extension, and truncation. An attempt is made to allocate space from the Extent Cache during file creation or extension. During file creation, if sufficient size is found, a hit is recorded. If the desired size is not found or an entry is forced to be split, an attempt is recorded. During file deletion, if the blocks were returned to the cache without the Extent Cache becoming too large, a hit is recorded. Otherwise a miss is recorded. Both hits and misses are counted as attempts. If quota checking is enabled, the Quota Cache is checked on file creation, deletion, extension, and truncation. If the desired entry (the identifier matching that of the requester) is found in the Quota Cache, a hit is recorded. Otherwise a miss is recorded. Both hits and misses are counted as attempts. The Bitmap Cache contains blocks from the storage bitmap file. This cache is accessed when the Extent Cache cannot satisfy requests for disk space. High rates indicate fragmented volumes. MON-56 MONITOR MONITOR FILE_SYSTEM_CACHE Data items in the FILE_SYSTEM_CACHE display correspond to SYSGEN ACP/XQP parameters as follows: FILE_SYSTEM_CACHE Item ACP/XQP Parameters Dir FCB ACP_SYSACC Dir Data ACP_DIRCACHE File Hdr ACP_HDRCACHE File ID ACP_FIDCACHE Extent ACP_EXTCACHE ACP_DINDXCACHE ACP_EXTLIMIT Quota ACP_QUOCACHE Bitmap ACP_MAPCACHE When you change the ACP/XQP cache parameters, remember to reboot the system to make the changes effective. For more information on these parameters, refer to the VAX/VMS System Generation Utility Reference Manual. MON-57 MONITOR MONITOR 10 MONITOR 10 The MONITOR IO command initiates monitoring of the IO class. FORMAT MONITOR 10 command qualifiers /qualifierl,...] class-name /ALL qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The IO class includes the following data items: * Direct I/O Rate—Rate of direct 1/O (for example, disk and tape) operations * Buffered I/O Rate—Rate of buffered 1/O (for example, terminal and line printer) operations ¢ Mailbox Write Rate—Rate of write-to-mailbox requests received by the system MON-58 * Window Turn Rate—Rate of file-map window misses * Log Name Translation Rate—Rate of logical name translations * File Open Rate—Rate at which files were opened * Page Fault Rate—Rate of occurrence of page faults for all working sets MONITOR MONITOR 10 ¢ Page Read Rate—Rate of pages read from disk as a result of page faults ¢ Page Read I/O Rate—Rate of read 1/O operations from disk as a result of page faults ¢ Page Write Rate—Rate of pages written to the page file ¢ Page Write /O Rate—Rate of write I/O operations to the page file * Inswap Rate—Rate at which working sets were read into memory from ® Free List Size—Number of pages on the free page list * Modified List Size—Number of pages on the modified page list the swap file EXAMPLE MONITOR> MONITOR /RECORD I0 VAX/VMS Monitor Utility 1/0 SYSTEM STATISTICS on node SAMPLE 16-MAY-1086 22:22:44 CUR AVE MIN Direct I/0 Rate 16.33 4.48 0.33 Buffered I/0 Rate 24.91 47.47 24.01 13.28 10.76 3.66 24.88 62.31 17.33 0.00 6.69 0.00 Mailbox Write Rate Window Turn Rate Log Name Translation Rate File Open Rate Page Fault Rate Page Read Rate Page Read 1/0 Rate Page Write Rate Page Write 1/0 Rate Inswap Rate Free List Size Modified List Size 0.00 1.66 1.68 12.20 2.65 0.00 0.00 3621.00 49.00 0.45 1.66 1.28 9.00 2.43 0.27 0.00 3604.09 73.36 MAX 0.00 0.33 0.33 0.00 0.00 0.00 0.00 3392.00 4.00 This example shows that the direct I/O rate is currently at its highest level since the MONITOR command was issued, and is significantly higher than the average rate. Termination of this command by CTRL/C and issuance of a MONITOR PROCESSES/TOPDIO command would show the top users of direct I/Os. Note that if I/O monitoring is begun at a later time, a new MONITOR request is defined. That is, it is not a continuation of the original request; the average, minimum, and maximum statistics are reinitialized. However, since the original request specified recording, that data can be played back for redisplay or summarization. MON-59 MONITOR MONITOR LOCK MONITOR LOCK The MONITOR LOCK command initiates monitoring of the LOCK class. FORMAT MONITOR LOCK command /qualifier],...] qualifiers class-name qual One or more qualifiers as described in the Command Qualifiers Section. /ALL Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The LOCK class includes the following data items: * New ENQ Rate—Rate of new lock (ENQ) requests (as opposed to conversions) * Converted ENQ Rate—Rate of lock (ENQ) conversion requests ¢ DEQ Rate—Rate of unlock (DEQ) requests * Blocking AST Rate—Rate of lock manager blocking ASTs delivered * ENQs Forced To Wait Rate—Rate of occurrence of locks that could not be granted immediately, thus having to wait * ENQs Not Queued Rate—Rate of occurrence of locks that could not be granted immediately but requested not to be queued, and thus received an error status instead MON-60 MONITOR MONITOR LOCK ¢ Deadlock Search Rate—Rate at which a deadlock search was performed ¢ Deadlock Find Rate—Rate at which a deadlock was found ¢ Total Locks—Total number of locks in the system * Total Resources—Total number of resources in the system EXAMPLE MONITOR> MONITOR /INPUT=LOCKSTATS.DAT/SUMMARY/NODISPLAY LOCK/AVERAGE NONITOR> $ TYPE MONITOR.SUM VAX/VMS Monitor Utility LOCK MANAGEMENT STATISTICS on node SAMPLE From: To: SUMMARY 0 P 5 2 |awer Converted ENQ Rate 1 e DEQ Rate Blocking AST Rate 3 [sasass Deadlock Search Rate Deadlock Find Rate Total Locks Total Resources PLAYBACK 16-MAY-1086 17:00:00 10 R New ENQ Rate ENQs Forced To Wait Rate ENQs Not Queued Rate 16-MAY-1986 08:00:00 16 20 BRI + I | } [} | i | | I | I | | I | I I | I | I 1 | | | 1 3 |wsnren 3 [wearar D SUMMARIZING | T | S I + This example shows the average use of the lock management subsystem during a typical workday, based on data that was previously recorded. MON-61 MONITOR MONITOR MODES MONITOR MODES The MONITOR MODES command initiates monitoring of the MODES class, which includes a data item for each mode of processor operation. FORMAT MONITOR MODES command /qualifierf,...] class-name /ALL qualifiers qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /[NOJCPU Selects the processor-specific form of display and summary output for the MODES class (for VAX-11/782 attached processor configurations). If the attached processor is not active, the qualifier has no effect. By default, processor-specific statistics are produced for VAX-11/782 systems for display output and single-file summaries. The qualifier has no effect on multifile summaries, which use only the non-processor-specific form. Note also that Pprocessor-specific statistics are not produced, even if explicitly requested, whenever SYSTEM is part of a list of classes. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. /[NOJPERCENT Controls whether statistics are expressed as percent values in display and summary output. The /PERCENT qualifier is applicable only to the DISK, MON-62 MONITOR MONITOR MODES MODES, SCS, and STATES classes. By default, statistics are not expressed as percent values. DESCRIPTION The following data items, included by the MODES class, can be displayed as percentages of all processor (CPU) time, or as rates of clock ticks (10 millisecond units) per second: * Interrupt Stack—Time spent on the interrupt stack * Kernel Mode—Time spent in kernel mode, but not on interrupt stack * Executive Mode—Time spent in executive mode * Supervisor Mode—Time spent in supervisor mode ® User Mode—Time spent in user mode executing VAX-11 instructions ¢ Compatibility Mode—Time spent executing compatibility mode instructions e Idle Time—Time spent executing the NULL process EXAMPLE MONITOR> MONITOR MODES /PERCENT VAX/VMS Monitor Utility TIME IN PROCESSOR MODES (%) 4meeeee 1 CuR | on node SAMPLE 15-MAY-1986 22:62:42 PR o4 26% 50% 8% P Interrupt Stack 4 1004 LR + Kernel Mode | 6 |*e | | | | | | | | | Executive Mode 2 | | | | | | | ] | i | | | I | | | | | | Supervisor Mode User Mode Compatibility Mode Idle Time | 1 72 [sxsarrenvanrenrenansesnren 1 16 [saeser This display shows that, over the last collection interval, the processor spent 72% of its time executing user code, 8% executing system code to service user requests in executive and kernel modes, and 4% processing interrupts on the interrupt stack; it was idle 16% of the time. Time spent executing VAX RMS code is included in executive mode time. Time spent executing DCL code is included in supervisor mode time. The majority of interrupt stack time is devoted to processing buffered I/O requests. If you omit the /PERCENT qualifier or specify /NOPERCENT, MONITOR displays mode times as rates of clock ticks per second, where a clock tick is 10 milliseconds. MON-63 MONITOR MONITOR MODES For a VAX-11/782 attached processor configuration, the MODES display consists of 14 items representing the 7 modes for each of the 2 processors. If you specify /NOCPU on such a system, only the 7 modes are displayed, where each mode includes the total time for both processors. On a single- processor system, the /CPU qualifier has no effect. Also, for a multifile summary, the qualifier has no effect (the non-processor-specific form is used). MON-64 MONITOR MONITOR PAGE MONITOR PAGE The MONITOR PAGE command initiates monitoring of the PAGE class. FORMAT MONITOR PAGE command qualifiers /qualifier],...] class-name /ALL qualifiers Specifies that a table of all available statistics (current, average, minimum, One or more qualifiers as described in the Command Qualifiers Section. and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. JAVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The PAGE class includes the following data items: ® Page Fault Rate—Rate of page faults for all working sets ¢ Page Read Rate—Rate of pages read from disk as a result of page faults ¢ Page Read I/O Rate—Rate of read I/O operations from disk as a result of page faults * Page Write Rate—Rate at which pages were written to the page file * Page Write I/O Rate—Rate of write I/O operations to the page file ® Free List Fault Rate—Rate at which pages were read from the free page list as a result of page faults MON-65 MONITOR MONITOR PAGE * Modified List Fault Rate—Rate of pages read from the modified page list as a result of page faults * Demand Zero Fault Rate—Rate at which zero-filled pages were allocated as a result of page faults * Global Valid Fault Rate—Rate of page faults for pages that are not in the process’s working set but are in physical memory and are indicated as valid pages in the systemwide global page tables * Writes In Progress Fault Rate—Rate of pages read that were in the process of being written back to disk, when faulted * System Fault Rate—Rate of page faults for pages in system space * Free List Size—Number of pages on the free page list ¢ Modified List Size—Number of pages on the modified page list _— EXAMPLE MONITOR> MONITOR PAGE VAX/VMS Momitor Usility PAGE MANAGEMENT STATISTICS on node SAMPLE 16-MAY-1086 22:22:44 CUR AVE MIN MAX Page Fault Rate 26.82 18.27 0.68 26.82 Page Vrite Raf Page Write I/0 Rate 0.00 0.00 0.00 0.00 Free List Fault Rate Page Page Rat 3.97 1.32 2.66 0.99 1.33 0.68 0.00 0.00 0.00 0.00 13.00 10.96 8.00 13.90 Demand Zero Fault Rate 4.63 2.65 0.68 4.63 ¥rt In Progress Fault Rate 0.00 Modified List Fault Rate Global Valid Fault Rate System Fault Rate Free List Size Modified List Size 5.62 1.32 2.31 3164.00 166.00 2.99 0.68 0.00 1.99 3176.00 131.00 0.33 0.00 3.97 1.32 5.62 1.32 0.00 0.00 3164.00 3188.00 1.68 107.00 2.31 166.00 This example shows that the current rate of pages read per read 1/O operation is approximately 3 per second (Page Read Rate divided by Page Read I/0 Rate). Note that while the page fault rate is currently at the highest point of the monitoring session, the vast majority of the pages are faulted from memory, not from disk. MON-66 MONITOR MONITOR POOL MONITOR POOL The MONITOR POOL command initiates monitoring of the POOL class, which measures space allocations in the nonpaged dynamic pool. FORMAT MONITOR POOL command qualifiers /qualifierl,...] class-name qualifiers /ALL One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. DESCRIPTION The POOL class includes the following data items: ® SRPs Available—Number of small request packets available in the SRP queue ® SRPs In Use—Number of small request packets available in use ¢ IRPs Available—Number of intermediate request packets available in the IRP queue * * IRPs In Use—Number of intermediate request packets available in use LRPs Available—Number of large request packets available in the LRP queue ® LRPs In Use—Number of large request packets in use MON-67 MONITOR MONITOR POOL * Dynamic Bytes Available—Number of dynamic bytes available * Dynamic Bytes In Use—Number of dynamic bytes in use * Holes In Pool—Unused blocks of contiguous space in the dynamically allocated portion of the pool ® Largest Block—Size in bytes of the largest block of unused space in the dynamically allocated portion of the pool * Smallest Block—Size in bytes of the smallest block of unused space in the dynamically allocated portion of the pool ® Blocks Less or Equal 32 Bytes—Blocks less than or equal to 32 bytes in size in the dynamically allocated portion of the pool EXAMPLE MONITOR> MONITOR/RECORD POOL/MINIMUM VAX/VMS Monitor Utility NONPAGED POOL STATISTICS on node SAMPLE 16-MAY-1986 10:22:39 oK e 50K e e ca- 76K s 450 | SRPs Available SRPs In Use 26K e 2010 IRPs Available IRPs In Use LRPs Available LRPs In Use [#snasasasnsrsass Dynanmic Bytes Available [#anasasasnonssasusaeasaranaranannonorans Dynamic Bytes In Use Holes 72 | In Pool Jossase Largest Block | Snallest Block Blocks or Eq 32 Bytes RECORDING In this example, only the minimum statistic is being displayed, but data is being recorded that could later be used to redisplay all statistics. Of particular interest in this display are the SRPs, IRPs, and LRPs Available items. If queue sizes for these preallocated packets drop to 0 at any time, extra overhead is incurred by the subsystems requiring these packets. Initial queue sizes of these items are determined by system parameters that can be adjusted at boot time. MON-68 MONITOR MONITOR PROCESSES MONITOR PROCESSES The MONITOR PROCESSES command initiates monitoring of the PROCESSES class, which displays information on all processes in the system. FORMAT MONITOR PROCESSES command /qualifierl,...] qualifiers One or more qualifiers as described in the Command Qualifiers Section. class-name qualifiers /TOPBIO Specifies that a bar-graph listing the top buffered I/O users be produced instead of the standard display and summary output. Values are expressed in units of buffered I/Os per second. /TOPCPU Specifies that a bar-graph listing the top CPU time users be produced instead of the standard display and summary output. Values are expressed in units of clock ticks (10 milliseconds) per second. /TOPDIO Specifies that a bar-graph listing the top direct I/O users be produced instead of the standard display and summary output. Values are expressed in units of direct I/Os per second. /TOPFAULT Specifies that a bar-graph listing the top page faulting processes be produced instead of the standard display and summary output. Values are expressed in units of page faults per second. DESCRIPTION As illustrated in the examples, the PROCESSES display (and summary) formats are different from those of all other classes. The PROCESSES display provides the following information: e PID—Process identification as assigned by the system, in hexadecimal ® STATE—Process’s scheduler state (See the description of the MONITOR STATES command for an explanation of the STATES codes.) ® PRI—Current (as opposed to base) priority of the process * NAME—Process name ¢ PAGES—Number of shareable pages and total number of pages currently in use by the process ¢ DIOCNT—Cumulative direct 1/O operations performed by the process since its creation; not displayed if the process is swapped out MON-69 MONITOR MONITOR PROCESSES * FAULTS—Cumulative page faults since the process was created; not displayed if process is swapped out * CPU TIME—Cumulative CPU time used by the process since its creation, in the format hours:minutes:seconds; not displayed if process is swapped out The top corners of the display contain the number of processes in the system and the time in days, hours, minutes, and seconds since the system was last booted. Processes that are swapped out are so noted. If more processes are in the system than can be displayed on the terminal screen at once, the full display occurs in screenfuls, presented one at a time with each display separated by an interval specified with the /VIEWING_ TIME qualifier. Note that the four /TOP bar-graph displays provide only the PID and process name of each of the top eight users. As with the other bar-graph displays, examples in the displays of top users are truncated to the nearest whole number. Up to eight processes with nonzero values are displayed. To be eligible for inclusion in the list of top users, a process must be present and swapped in at the beginning and end of the display interval. This eligibility requirement also applies to the beginning and ending of the entire period covered by a summary. Note that only one of the displays of top users or the regular PROCESSES display can be selected in a single MONITOR request. EXAMPLES 0 MONITOR> MONITOR/INPUT=PROCS.DAT/INTERVAL=6 PROCESSES Process Count: 20 VAX/VMS Momitor Utility Uptime: 1 23:26:10 PROCESSES on node SAMPLE 15-MAY-1986 12:39:00 PID STATE PRI NAME PAGES 00000080 COM O NULL 0/0 00000081 HIB 00000102 LEFO 16 SWAPPER 4 SAMPLE1001 0/0 87/232 00000103 COM 4 SAMPLE1101 16/100 00000084 HIB 8 ERRFMT 64/174 00000086 LEF 00000087 HIB 8 OPCOM 9 JOB_CONTROL 73/272 67/203 00000088 HIB 8 CONFIGURE 43/206 0000008A HIB 6 SYMBIONT_0001 6/66 0000008B HIB 8 JNLACP 76/680 0000008C HIB 0000008D HIB 0000008E HIB 8 NETACP 5 EVL 9 REMACP 6/9654 7/66 6/64 00000112 COM 4 SAMPLE1601 46/111 0000011E CUR 9 SMITH 89/208 DIOCNT FAULTS CPU TIME PLAYBACK This example illustrates a PROCESSES display generated from the input file PROCS.DAT. One line is displayed for each process in the system. Note that this display shows current values only—average, minimum, and maximum statistics are not available. Also note that for swapped-out processes, the words SWAPPED OUT replace the three rightmost items, because those items are not available for swapped-out processes. Since this example is a playback request, the system uptime displayed is that of the system at the time the MONITOR data was recorded. MON-70 MONITOR MONITOR PROCESSES Nondisplayable characters in process names are represented by periods. B MONITOR> MONITOR/TOPDIO/INPUT=PROCS.DAT PROCESSES VAX/VMS Monitor Utility TOP DIRECT I/0 RATE PROCESSES on node SAMPLE 16-MAY-1986 16:13:38 [ 2% 50 7 D 000000C7 SAMPLEOS01 25 [rewesanss 00000112 SAMPLE1601 17 [aneane 00000102 SAMPLE1001 14 00000103 SAMPLE1101 12 NULL 12 0000011E SMITH 4 0000008C NETACP P\ | | I | 100 e+ | PLAYBACK This example shows that the process SAMPLE0901, with a rate of 25 per second, was the top consumer of direct I/Os during the most recent interval between displays. MON-71 MONITOR MONITOR SCS MONITOR SCS The MONITOR SCS command initiates monitoring of the SCS (System Communication Services) class. FORMAT MONITOR SCS command qualifiers /qualifier,...] One or more qualifiers as described in the Command Qualifiers Section. class-name qualifiers /ALL Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /ITEM=(keyword],...]) Selects one or more data items for inclusion in display and summary output. If you specify two or more keywords, enclose them in parentheses and separate them with commas. When the /ITEM qualifier is omitted, the default is /ITEM=KB_MAP. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. /[NOJPERCENT Controls whether statistics are expressed as percent values in display and summary output. The /PERCENT qualifier is applicable only to the DISK, MODES, SCS, and STATES classes. By default, statistics are not expressed as percent values. MON-72 MONITOR MONITOR SCS /ITEM Qualifier Keywords ALL Specifies that statistics on all data items collected for the disks are displayed on successive screens. BUFFER_DESCRIPTOR Specifies that statistics on the queued-for-buffer-descriptor (on the local node) rate are displayed for each node. D_DISCARD Specifies that datagram discard rate statistics are displayed for each node. D_RECEIVE Specifies that datagram receive rate statistics are displayed for each node. D_SEND Specifies that datagram send rate statistics are displayed for each node. KB_MAP Specifies that kilobyte map rate statistics are displayed for each node. KB_REQUEST Specifies that kilobyte request (via request datas) rate statistics are displayed for each node. KB_SEND Specifies that kilobyte send (via send datas) rate statistics are displayed for each node. M_RECEIVE Specifies that message receive rate statistics are displayed for each node. M_SEND Specifies that message send rate statistics are displayed for each node. REQUEST_DATA Specifies that request data (initiated on the local node) rate statistics are displayed for each node. SEND_CREDIT Specifies that queued-for-send-credit (on the local node) rate statistics are displayed for each node. SEND_DATA Specifies that send data (initiated on the local node) rate statistics are displayed for each node. DESCRIPTION The SCS class is a component class. Data items for this class are collected for each node in the cluster. The SCS class consists of the following data items: * ® Datagram Send Rate—Rate at which datagrams are sent to another node. Datagram Receive Rate—Rate at which datagrams are received from another node. MON-73 MONITOR MONITOR SCS ® Datagram Discard Rate—Rate at which datagrams are discarded by the Cluster Interconnect (CI) port driver. Datagrams are discarded by the CI port driver when a receive buffer is not available. Datagrams can also be discarded by the CI port hardware, but this is not reflected by the Datagram Discard Rate. ® Message Send Rate—Rate at which sequenced messages are sent to another node. Sequenced messages are exchanged between nodes to communicate with Mass Storage Control Protocol (MSCP) disks and the lock manager. * Message Receive Rate—Rate at which sequenced messages are received from another node. Sequenced messages are exchanged between nodes to communicate with MSCP disks and the lock manager. * Send Data Rate—Rate at which block send datas are initiated on the local node. * * Kbytes Send Rate—Rate at which kilobytes are sent, as a result of send datas initiated on the local node. Request Data Rate—Rate at which request datas are initiated on the local node. * Kbytes Request Rate—Rate at which kilobytes are received, as a result of request datas initiated on the local node. ¢ Kbytes Map Rate—Rate at which kilobytes are mapped for block transfers. This is a rough measure of the data transfer rate between the local node and a remote node. Before any transfer can take place, a buffer must be mapped. The size of the accumulated buffers that were mapped is displayed by the Kbytes Map Rate. If request datas or send datas are initiated on the local or a remote node, then the Kbytes Map Rate reflects the number of kilobytes actually transferred between the two nodes. ® Send Credit Queued Rate—Rate at which connections are queued for a send credit. A connection is queued for a send credit whenever all of the buffers that were allocated by the remote node have been used. * Buffer Descriptor Queued Rate—Rate at which connections are queued for a buffer descriptor. A connection is queued for a buffer descriptor whenever all of the buffer descriptors have been allocated by the local node. You can increase the number of buffer descriptors allocated on the local system by adjusting the system parameter SCSBUFFCNT. R EXAMPLE MONITOR> MONITOR SCS VAX/VMS Monitor Utility SCS STATISTICS on node CURLEY 16-MAY-1986 10:21:48 Kbytes Map Rate CURLEY MOE LARRY SHEMP CUR 0.00 0.00 0.00 65.64 AVE 0.00 0.00 0.00 3.81 MIN 0.00 0.00 0.00 1.98 MAX 0.00 0.00 0.00 5.64 The command in this example requests that kilobyte map rate statistics collected for SCS be displayed for each node in the VAXcluster. The display shows block transfer map activity between the node CURLEY and MON-74 MONITOR MONITOR SCS the Hierarchical Storage Controller (HSC) SHEMP. Note that each node in the VAXcluster is identified by its SCS node name. MON-75 MONITOR MONITOR STATES MONITOR STATES The MONITOR STATES command initiates monitoring of the PROCESS STATES class, which shows the number of processes in each of the 14 scheduler states. FORMAT MONITOR STATES command /qualifier],...] class-name /ALL qualifiers qualifiers One or more qualifiers as described in the Command Qualifiers Section. Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar-graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar-graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar-graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar-graph for display and summary output. /[NOJPERCENT Controls whether statistics are expressed as percent values in display and summary output. The /PERCENT qualifier is applicable only to the DISK, MODES, SCS, and STATES class names. By default, statistics are not expressed as percent values. MON--76 MONITOR MONITOR STATES DESCRIPTION The STATES class shows the number of processes in each of the 14 scheduler states as follows: * * Collided Page Wait (COLPG)—Waiting for a faulted page in transition Mutex & Miscellaneous Resource Wait (MWAIT)—Waiting for the availability of a mutual exclusion semaphore or a dynamic resource. Following is a summary of Mutex and Miscellaneous Resource Wait states and identifying codes, as they appear in the PROCESSES class display. Resource Wait Code Reason for Wait RWAST AST wait (wait for system or special kernel AST) RWBRK Breakthrough (wait for broadcast message) RWCLU Cluster state transition wait RWIMG Image activation lock RWLCK Lock data base RWMBX Mailbox full RWMPB Modified page writer busy RWMPE Modified page list empty RWNPG Nonpaged dynamic memory RWPAG Paged dynamic memory RWPGF Page file full RWQUO Job quota RWSCS System Communication Services wait RWSWP Swap file space ¢ Common Event Flag Wait (CEF)—Waiting for some combination of event flags to be set in a common event block ® Page Fault Wait (PFW)—Waiting for a page to be read as a result of a page fault; resident processes * ® Local Event Flag Wait (LEF)—Waiting for one or more local event flags to be posted; resident processes Local Event Flag (Outswapped) (LEFO)—Waiting for one or more local event flags to be posted; outswapped processes * Hibernate (HIB)—Hibernating, or process has executed a hibernate request; resident processes * Hibernate (Outswapped) (HIBO)—Hibernating, or process has executed a hibernate request; outswapped processes ® Suspended (SUSP)—Process has executed a suspend request; resident processes * Suspended (Outswapped) (SUSPO)—Process has executed a suspend request; outswapped processes ® Free Page Wait (FPW)—Waiting for a free page of memory MON-77 MONITOR MONITOR STATES ® Compute (COM)—Ready to use the processor; resident processes * Compute (Outswapped) (COMO)—Ready to use the processor; outswapped processes ® Current Process (CUR)—Using the processor The data items can also be displayed as percentages of all processes. Note that the Current Process is always the process running MONITOR, since MONITOR is running when each measurement is made. Local Event Flag Wait (Outswapped) processes normally belong to interactive users who have been prompted but have not responded, although they might be processes waiting for disk I/O on a crowded system. A state of Compute (Outswapped) for any process indicates a very crowded system. For performance reasons, MONITOR does not synchronize the scanning of process state data structures with operating system use of those structures. It is therefore unlikely that MONITOR will display certain anomalous state indications. EXAMPLE $ MONITOR/INPUT/SUMMARY/NODISPLAY - $_/BEGINNING=15-MAY-108! 3. $_ /ENDING=15-MAY-1085: 1 :00 - STATES/PERCENT/ALL $ TYPE MONITOR.SUM VAX/VMS Monitor Utility PROCESS STATES (%) Suspended Suspended (Outswapped) Free Page Wait Compute Compute (Outswapped) Current Process PLAYBACK 43 g 868 sco0 0.0 TS Hibernate (Outswapped) MINY, NNWooOOA RraBWOOO Hibernate 15-MAY-1986 14:00:00 suboocoh Local Evt Flg (Outswapped) 15-MAY-1986 13:00:00 ro~NOOO Local Event Flag Wait howooow NoNmoOOH Page Fault Wait s Connon Event Flag Wait X ©oooo0o0o0 roNwoOO Collided Page Wait Mutex & Misc Resource Wait roNocoon 8olroocox CUR! To: wowoooo From: SUMMARY roroooo TM Bokooo on node SAMPLE SUMMARIZING The commands in this example generate and display a PROCESS STATES summary. Note that since use of the RETURN key is not permitted within a single MONITOR command following the MONITOR> prompt, the MONITOR command is issued at DCL level. The summary shows that on the average 14.1% of processes were swapped out for the summarized period. Note that the summary was requested for data covering only the hour between 1 P.M. and 2 P.M,, although the input file could have contained data covering a longer period. MON-78 MONITOR MONITOR SYSTEM MONITOR SYSTEM The MONITOR SYSTEM command initiates monitoring of the SYSTEM statistics class, which shows several of the most important items from other classes. FORMAT MONITOR SYSTEM command /qualifier],...] qualifiers class-name qualifiers One or more qualifiers as described in the Command Qualifiers Section. /ALL Specifies that a table of all available statistics (current, average, minimum, and maximum) is to be included in the display and summary output. For summary output, this qualifier is the default for all classes; otherwise, it is the default for all classes except CLUSTER, MODES, PROCESSES, STATES, and SYSTEM. /AVERAGE Selects average statistics to be displayed in a bar graph for display and summary output. /CURRENT Selects current statistics to be displayed in a bar graph for display and summary output. The /CURRENT qualifier is the default for the CLUSTER, MODES, STATES, and SYSTEM classes. /MAXIMUM Selects maximum statistics to be displayed in a bar graph for display and summary output. /MINIMUM Selects minimum statistics to be displayed in a bar graph for display and summary output. DESCRIPTION Because the SYSTEM class collects the most significant performance statistics from other classes in a single display, it is particularly useful to system managers and other users seeking a general overview of system activity. The SYSTEM class includes the following data items: Interrupt Stack Kernel Mode Executive Mode Supervisor Mode User Mode Compatibility Mode Idle Time Process Count Page Fault Rate MON-79 MONITOR MONITOR SYSTEM Page Read 1/0 Rate Free List Size Modified List Size Direct 1/0 Rate Buffered 1/0 Rate Two display formats are provided, depending on the class-name qualifier specified: ® A tabular style format for the /ALL qualifier ® A bar-graph style format for the /AVERAGE, /CURRENT, /MAXIMUM, and /MINIMUM qualifiers Note that the bar-graph version of the SYSTEM class (shown in Example 2) contains somewhat different data from the tabular version: ® All of the CPU processor modes except Idle Time are summed in the CPU Busy segment. * In the Page Fault segment, the page read I/O rate is indicated by a vertical bar. The bar provides a visual estimate of the proportion of the total page fault rate which caused read 1/O operations (the “hard” fault rate). The “hard” fault rate appears to the left of the bar. * Four segments show the processes which are currently (since the last screen update) the top consumers of CPU, page faults, direct I/Os, and buffered I/Os. Note that the NULL process is never displayed, even if it is a top consumer. Any process that MONITOR designates as a top user process must be swapped in at the beginning and end of the display interval, or at the beginning and ending of the entire period covered by a summary. When the upper bar graph (top user) and the corresponding lower bar graph (overall system measure) are tracking the same statistic for the same interval (as in Example 2), it is reasonable to compare the two graphs. This will be the case whenever — SYSTEM is the only class being monitored (no other class names have been specified with the MONITOR command). — The CURRENT statistic is specified. — The /INTERVAL and /VIEWING_TIME values are equal. Otherwise, you should exercise care in making such comparisons for the following reasons: — The top user statistic is always CURRENT, while the overall system measure statistic may be CURRENT, AVERAGE, MAXIMUM, or MINIMUM. — Rates for top users are calculated based on the interval between two successive screen displays, while overall system rates are based on the collection interval. These two interval values can be different whenever one or more classes are being monitored with the SYSTEM class, or when /INTERVAL and /VIEWING_TIME values differ. * While other upper boundary figures for the SYSTEM class bar graphs are constants, the figures for Free List Size and Modified List Size are derived from the physical memory configuration and system parameters of individual systems. The upper boundary figure for the Free List is the number of pages available after deducting the pages permanently allocated MON-80 MONITOR MONITOR SYSTEM to the VAX/VMS operating system. This figure, sometimes referred to as “balance set memory,” is the number of pages that can be allocated to processes, the Free List, and the Modified List. The upper boundary figure for the Modified List is the value of the MPW_HILIMIT system parameter. Note that both upper boundary figures are calculated when the MONITOR request is initiated and do not change thereafter. L] EXAMPLES MONITOR> MONITOR SYSTEM/ALL VAX/VMS Monitor Utility SYSTEM STATISTICS on node SAMPLE 16-MAY-1086 12:43:28 CUR AVE MIN MAX Interrupt Stack 0.33 0.33 0.33 0.33 Kernel Mode 0.16 0.16 0.16 0.16 99.00 98.67 99.00 99.00 14.00 14.00 14.00 14.00 0.00 0.00 0.00 0.00 0.00 0.50 0.00 Executive Mode Supervisor Mode User Node Compatibility Mode Idle Time s Count Page Fault Rate 0.00 0.00 0.49 0.00 0.33 Page Read 1/0 Rate Free List Size 0.00 0.00 0.50 0.00 0.33 0.00 0.00 0.50 0.00 0.33 0.33 0.00 4266.00 4266.00 4266.00 4265.00 0.00 0.00 0.00 0.00 106.00 Moditied List Size Direct 1/0 Rate Buffered 1/0 Rate 106.00 0.16 106.00 0.18 106.00 0.16 0.16 This example shows the tabular style format for the SYSTEM display. MONITOR> MONITOR SYSTEM SAMPLE VAX/VMS Monitor Utility Statistic: CURRENT Node: SYSTEM STATISTICS 16-MAY-1986 12:38:48 LEF. HIB: -+ 11 HIBO: 100 rooo Process Stat -+ Cur Top: ASSEM_LIBRTL (34) + Page Fault Rate (25) MEMORY -+ I lax]one O + Modified List Size (146) + Cur Top: ASSEM_SYS (11) + Direct I1/0 Rate (165) [ryeyen 1/0 -+ + Buffered I1/0 Rate (0) -+ | This example shows the bar-graph style format for the SYSTEM display. MON-81 MONITOR SET DEFAULT SET DEFAULT The SET DEFAULT command sets command qualifier and classname parameter defaults for the MONITOR command. Each time you issue the SET DEFAULT command, only the command qualifiers you specify explicitly are set. However, each new SET DEFAULT command replaces the entire set of class-name parameters. FORMAT SET DEFAULT classname],...] command classnamel,...] parameters Specifies one or more class names. command /qualifier],...] class-name /qualifier],...] qualifiers qualifiers One or more qualifiers as described in the Command Qualifiers Section. One or more class-name qualifiers (see Table MON-1). —_— DESCRIPTION Command and class-name qualifiers are identical to those for the MONITOR classname commands. S — EXAMPLE MONITOR> SET DEFAULT /INTERVAL=10 PAGE/AVERAGE+POOL/MAXIMUM The command in this example selects PAGE and POOL as the default classes for the MONITOR classname command, and specifies an interval of 10 seconds for the statistics display. The command specifies that AVERAGE statistics be displayed for the PAGE class and that MAXIMUM statistics be displayed for the POOL class. After establishing these defaults, you can issue the MONITOR command without any qualifiers or parameters to display the specified results. MON-82 MONITOR SHOW DEFAULT SHOW DEFAULT The SHOW DEFAULT command displays the defaults established by the SET DEFAULT command. FORMAT SHOW DEFAULT command None. parameters command qualifiers None. DESCRIPTION You can use the SHOW DEFAULT command to verify the defaults you have set with the SET DEFAULT command. EXAMPLE MONITOR> SHOW DEFAULT /BEGINNING = current time /ENDING = indefinite /INTERVAL =10 /VIEWING_TIME = 10 /NOINPUT /NORECORD /DISPLAY = SYS$OUTPUT: . ; /NOSUMMARY /NOCOMMENT Classes: PAGE/AVERAGE POOL/MAXIMUM The command in this example displays the defaults specified by the previous SET DEFAULT command. MON-83 MONITOR Examples _ ——————————— LIVE DISPLAY MONITORING The following examples illustrate the live display monitoring mode of EXAMPLES operation. Use this mode when you want to examine the activity of a running system, either on a routine basis, or as part of an installation checkout, tuning, or trouble-shooting exercise. No historical record of output is kept. D MONITOR PROCESSES/TOPCPU This command displays a bar graph showing the eight processes that were the top consumers of CPU time during the period between displays. It also displays the amount of CPU time each of these processes used. B $ MONITOR/DISPLAY=POOL.LOG POOL MONITOR display output can be routed to any supported terminal device or to a disk file. The command in this example writes MONITOR's display of nonpaged pool statistics to the file POOL.LOG. This file could then be printed out on a hardcopy device. B My_cLASSES :== - _8$ "DECNET+FCP+I0+LOCK+MODES+PAGE+POOL +PROCESSES+STATES" $ MONITOR/NODE=(CURLEY,LARRY)/INTERVAL=20/VIEWING_TIME=8 'MY_CLASSES' You may find it convenient to establish DCL symbols for frequently used combinations of class names, as in this example. Here, the MONITOR command collects selected classes of data for nodes CURLEY and LARRY every 20 seconds. Every 8 seconds, the most recently collected data for one of the classes is displayed. The ordering of the classes for display is predetermined by MONITOR. =_—— LIVE RECORDING EXAMPLE ——————————— The following example illustrates the live recording mode of operation. Use live recording whenever you need to capture MONITOR data for future use. Possible uses include the following: ¢ Installation checkout, tuning, trouble-shooting—that is, all the uses listed above for live display monitoring. Choose recording over display whenever you would like to capture more classes than you could reasonably watch at a terminal, whenever a terminal is not available, or whenever you need to gather data about the system but cannot devote your time to the terminal until later. ¢ Routine performance data gathering for long-term analysis. MONITOR data can be recorded on a routine basis and summarized to gather data about system resource utilization over long periods of time. $ MONITOR/NODE=(LARRY ,MOE)/NODISPLAY/RECORD MODES+STATES This command records data on the time spent in each of the processor modes and on the number of processes in each of the scheduler states for nodes LARRY and MOE. It does not, however, display this information. MON-84 MONITOR Examples CONCURRENT DISPLAY AND RECORDING EXAMPLES The following examples illustrate the concurrent display and recording mode of operation. Use this mode whenever you want to retain performance data and watch as it is being collected. 0 # MONITOR/RECORD FCP/CURRENT,POOL/MINIMUM,FCP/AVERAGE This command collects and records file system ACP data and nonpaged dynamic pool data every three seconds. It also displays, in bar-graph form, average FCP statistics and minimum POOL statistics. The display alternates between the two graphs every three seconds. Note that the FCP class was requested twice; while this is a misuse of the MONITOR command, it is not flagged as an error. Since MONITOR allows only one statistic qualifier per class name, only the later specification (/AVERAGE) is used. Current statistics can be obtained in a subsequent playback request. Another method of performing concurrent display and recording uses two separate processes, one for recording and one for display. The display process plays back the recording file as it is being written (possible because MONITOR allows shared read access to the recording file). The next example shows how one process (Process A) can perform recording, while another process (Process B) plays back the file to obtain a summary. Process A issues the following command: B $ MONITOR/RECORD=SYSSMANAGER: ARCHIVE.DAT: -8 /INTERVAL=300/NODISPLAY ALL_CLASSES) Process A uses this command to archive data for all classes once every three minutes. System managers may find it convenient to place a similar command in their system startup command file. Process B now issues the command: # MONITOR/INPUT=SYS$MANAGER: ARCHIVE.DAT: —$ /NODISPLAY/SUMMARY/BEGINNING="-1" PAGE,I0) Using this command, Process B obtains a summary of page and 1/O activity that occurred during the previous hour, perhaps as part of an investigation of a reported performance problem. Note that since Process A, while recording, executes an RMS flush operation every five minutes, up to five minutes of the most recently collected data will not be available to process B. The time between flush operations can be specified explicitly with the /FLUSH._ INTERVAL qualifier. Note also that Process B must have read access to the recording file. MON-85 MONITOR Examples _—_ PLAYBACK EXAMPLES The following examples illustrate the playback mode of operation. Use playback of a recording file to obtain terminal output and/or summary reports of all or just a subset of collected data. Data can be subsetted by class or time segment. For example, if several classes of data have been collected for an entire day, you can examine or summarize the data on any one or more of the classes during any time period in that day. You can also display or summarize data with a different interval than that at which it was recorded. The actual amount of time between displays of screen images is controlled with the /VIEWING_TIME qualifier. f ¢ MONITOR/RECORD/INTERVAL=6 10 $ MONITOR/INPUT 10 These commands produce system 1/O statistics. The first command gathers and displays data every five seconds, beginning when the command is issued and ending when the user issues a CTRL/C. In addition, it records binary data in the default output file MONITOR.DAT. The second command plays back the 1/0 statistics display, using the data in MONITOR.DAT for input. The default viewing time for the playback is three seconds, but each screen display represents five seconds of monitored 1/0 statistics. B $ MONITOR/RECORD/NODISPLAY $ /BEGINNING=08:00:00 -8 /ENDING=16:00:00 _$ /INTERVAL=120 POOL $ MONITOR/INPUT/DISPLAY=HOURLY.LOG _$ /INTERVAL=3600 POOL This sequence of commands illustrates the recording of data with a relatively small interval and playback with a relatively large interval. This is useful for producing average, minimum, and maximum statistics that cover a wide range of time, but have greater precision than they would have if they had been gathered using the larger interval. The first command records data on space allocation in the nonpaged dynamic pool for the indicated eight-hour period, using an interval of two minutes. The second plays the data back with an interval of one hour, storing display output in the file HOURLY.LOG. This file can then be typed or printed to show the cumulative pool utilization at each hour throughout the eight-hour period. E ¢ MONITOR/INPUT/NODISPLAY/SUMMARY=DAILY.LOG POOL/AVERAGE This command uses the recording file created in the previous example to produce a one-page summary report file showing the average statistics for the indicated eight-hour period. The summary report has the same format as a screen display, which in this case is a bar graph. MON-86 MONITOR Examples __ REMOTE PLAYBACK EXAMPLE You can collect MONITOR data from any system to which your system has a DECnet connection. You can then display the data “live” on your local system. To implement the procedure, follow these steps: 1 In the default DECNET directory on each remote system, create a file, MONITOR.COM, similar to the following: $ $ ! * Enable MONITOR remote playback * s ! 3 MONITOR /NODISPLAY/RECORD=SYS$NET ALL_CLASSES 2 On your local system, define a logical name, as follows, for the remote system from which you wish to collect data: $ DEFINE remotenodename_mon "node: :""task=monitor""" You may want to define, in a login command procedure, a series of such logical names for the various systems you wish to access. To display the remote MONITOR data as it is being collected, issue a command of the form: $ MONITOR/INPUT=remotenodename_mon classnames You may also place MONITOR.COM files in directories other than the default DECNET directory and use access control strings and/or proxy accounts to invoke these command files remotely. When you invoke MONITOR on your local system, a process is created on the remote system which executes the MONITOR.COM command file. The remote system will therefore experience some associated CPU and DECnet overhead. You can regulate the overhead in the MONITOR.COM file through the use of the /INTERVAL qualifier and the list of class names. _— ( RERECORDING EXAMPLE The following example illustrates the rerecording mode of operation. Rerecording is a combination of playback and recording. It can be useful for data reduction of recording files. When you play back an existing recording file, all MONITOR options are available to you; thus, you can choose to record a subset of the recorded classes and/or a subset of the recorded time segment and even a larger interval value. All these techniques produce a new, smaller recording file, at the expense of some of the recorded data. A larger interval value reduces the volume of the collected data, so that displays and summary output produced from the newer recorded file will be less precise. Note that average rate values will not be affected in this case, but average level values will be less precise (since the sample size is reduced), as will maximum and minimum values. $ SUBMIT MONREC.COM MONREC.COM contains the following commands: $ MONITOR/NODISPLAY/RECORD/INTERVAL=60 /BEGINNING=8:00/ENDING=16:00 DECNET,LOCK $ MONITOR/INPUT/NODISPLAY/RECORD DECNET MON-87 MONITOR Examples The first command runs in a batch job, recording DECnet-VAX and lock management data once per minute, between the hours of 8 A.M. and 4 P.M. The second command, which is issued after the first command completes, rerecords the data by creating a new version of the MONITOR.DAT file, containing only the DECnet-VAX data. s s -_—_b§ VAXCLUSTER MULTIFILE SUMMARY This section provides an example of a procedure that you, as cluster manager, EXAMPLE might use to create multifile summaries in a VAXcluster environment (cluster summaries). You will probably want to write your own command procedures to suit conditions at your site. To facilitate generation of cluster summaries, and to perform other “automatic” MONITOR functions, you can create and execute command files similar to those illustrated in the accompanying procedure (and supplied in SYSJEXAMPLES). Note that you must define the logical names SYS$MONITOR and MON$ARCHIVE in SYSTARTUP.COM before executing any of the command files. To produce a cluster summary, create and execute command files similar to the following: Fe R R 1 MONITOR.COM—Archive recording file and summary file from previous boot; initiate continuous recording for current boot SET VERIFY ' ' ' MONITOR. COM ' This command file is to be placed in ! process via SUBMON.COM. ' | ! ' ' ' called SYS$MONITOR and submitted at en startup tim ble directory For each node, MONITOR.COM cre detached in SYSSMONITOR, a MONITOR recording file that is updated throughout the 1ife of the boot. It also creates, in MONSARCHIVE, a summary file from the recording file of the previous boot, along with a copy of that recording file. Logical name definitions for both cluster-accessible directories, SYSSMONITOR and MONSARCHIVE, should be included in SYSTARTUP.COM. ' s ET DEF SYS$MONITOR SET NOON PURGE MONITOR.LOG/KEEP:2 1 ! ! Compute executing nede name and recording and sumnmary file names (incorporating node name and date). ' NODE = F$GETSYI("NODENANE") SEP = " IF NODE .NES. "* THEN SEP = "_" DAY = F$EXTRACT (0,2,F$TIMEQ)) IF F$ERTRACT(O,1, .EQS. " " THEN DAY = F$EXTRACT(1,1,DAY) MONTH = F$EXTRACT(3,3,FSTIME()) ARCHFILNAM = "NONS$ARCHIVE: *+NODE+SEP+"NON" +DAY+MONTH RECFIL = NODE+SEP+"MON.DAT® SUMFIL = ARCHFILNAM+".SUM" Check for existence of recording file from previous boot and skip sumnary if not present. MON-88 MONITOR PRI OPEN/READ/ERROR=NORECFIL RECORDING 'RECFIL' CLOSE RECORDING ! ! ! Generate summary file from previous boot. ! MON: ITOR /INPUT='RECFIL' /NODISPLAY /SUMMARY='SUMFIL' ALL_CLASSES+MODE/ALL+STATES/ALL+SCS/ITEM=ALL+SYSTEM/ALL +DISK/ITEM=ALL ! ' ! Compute subject string and mail summary file to cluster manager. ' ! > 1 U Examples B=" R R ! ! Rk R R ' T MONITOR Summary " SUB = A+NODE+B+FSTINE()+A MAIL/SUBJECT='SUB' 'SUMFIL' CLUSTER_MANAGER Archive recording file and delete it from SYS$MONITOR. ' COPY 'RECFIL' 'ARCHFILNAM'.DAT DELETE 'RECFIL';+ ' R NORECFIL: SET PROCESS/PRICRITY=15 Begin recording for this boot. The specified /INTERVAL value is adequate for long-tern summaries; you may require a smaller val to get reasonable ive" playback summaries R of more disk space for the recording file). NITOR /INTERVAL=300 /NODISPLAY /RECORD='RECFIL' ALL_CLASSES End of MONITOR.COM 2 Crssnccnnnennn (at the expense SUBMON.COM—Submit MONITOR.COM as a detached process from SYSTARTUP.COM to initiate continuous recording for the current boot. SET VERIFY SUBMON . COM This comnand file is to be placed in a cluster-accessible directory called SYS$HONITOR. At system startup time, for executed by SYSTARTUP.COM, each node, it is following logical neme definitions for the cluster-accessible directories SYS$MONITOR and MON$ARCHIVE. Submit detached MONITOR proce RUN to do continuous recording. SYS$SYSTEM:LOGINOUT.EXE - /urc=(1,4] G /INPUT=SYSSHONITOR: NONITOR.COM /OUTPUT=SYS$MONITOR : MONITOR. LOG - /ERROR=SYS$MONITOR:MONITOR.LOG wonn /PROCESS_NAME="Monitor" /WORKING_SET=100 /MAXINUM_WORKING_SET=100 /EXTENT=612/NOSWap - End of SUBMON.COM MON-89 MONITOR Examples BB R BB BB BN S| ET VERIFY RRNRN RN R BN N 3 MONSUM.COM—Generate daily and “prime-time” cluster summaries: SET DEF SYS$MONITOR 1 MONSUM. COM 1 ! This command file is to be placed in a cluster-accessible directory ! called SYSSNONITOR and executed at the comvenience of the cluster both 24-hour and "prime time" cluster nmanager. The file generates ' ' suznaries and resubmits itself to run each day at midnight. ! ! SET NOON ' Compute file specification for MONSUM.COM and resubmit the file. | ! FILE = F$ENVIRONMENT("PROCEDURE") FILE = FS$PARSE(FILE, , , "DEVICE")+F$PARSE(FILE, , , "DIRECTORY") +F$PARSE(FILE, , , "NAME") SUBMIT 'FILE' /AFTER=TOMORROW /NOPRINT ' Ges ! o 24-hour cluster summary. ' ! MONITOR/INPUT=(SYS$MONITOR: #MON+ .DAT; # MON$ARCHIVE: +MONs .DAT;#) - SUM /NODISPLAY/SUMMARY=MONSUM. ALL_CLASSES+DISK/ITEM=ALL+SCS/ITEM=ALLcnnnnesenneen /BEGIN="YESTERDAY+0:0:0.00" /END="TODAY+0:0:0.00" /BY_NODE ' Mail 24-hour summary file to cluster manager and delete the file from SYS$MONITOR. MAIL/SUBJECT="Daily Monitor Clusterwide Summary” MONSUM.SUM CLUSTER_MANAGER DELETE MONSUM.SUM; % ! ! Generate prime-time cluster summary. ' ! #MON+ .DA' MONITOR/INPUT=(SYS$MONITOR: . SUM /NODISPLAY/SUMMARY=MONSUM VE: #MON+ .DAT;#) ,MONSARCHI wnnsnasnnnn ALL_CLASSES*+DISK/ITEM=ALL+5CS/ITEM=ALL/BEGIN="YESTERDAY+9:0:0.00" /END="YESTERDAY+18:0: .00" /BY_NODE ' Mail prime-time summary file to cluster manager and delete the file from SYS$MONITOR. MAIL/SUBJECT="Prime-Time Monitor Clusterwide Summary" MONSUM.SUM CLUSTER_MANAGER DELETE MONSUM.SUM; ' ! End of NONSUM.COM ! Remember that the summary data may be extensive; you should therefore plan to print out the summary files. MON-90 MONITOR Supplemental MONITOR Information A Supplemental MONITOR Information This section describes the MONITOR recording file record formats. A.1 The MONITOR Recording File — Binary performance data is written into the MONITOR recording file when a MONITOR request indicates recording. A record is written to this file once for each requested class per interval; the record contains a predefined set of data for each of the requested performance classes. The recording file is created when a MONITOR request is initiated, and is closed when the request terminates. The MONITOR recording file may be used as a source file to format and display the data on a terminal, to create a summary file, or to record a new recording file with different characteristics. Note: The record formats described in this section are subject to change without notice at any future major VAX/VMS release. The MONITOR recording file is a VAX RMS sequential file with variablelength records. Each record in the file begins with a one-byte type field. The remaining fields are different in length and format for each record type. There are three categories of record types: * Customer control record * DIGITAL control record ® Class record Customer control records may appear anywhere in the recording file. They are not generated by MONITOR and are ignored by MONITOR when it reads the file. The first records in the MONITOR recording file, excluding customer control records, are DIGITAL control records. There are currently three types of DIGITAL control records: the file header record, the system information record, and the node transition records—one for each node. The class records, which contain data on requested performance classes, follow the DIGITAL control records. The class record is generally written once per interval for each class being recorded. An exception to this rule is the case where several class records are required to contain data for a single class over a single interval. This case can occur for the PROCESSES class when there are too many processes to be accommodated by the maximum record size. Unique numbers are assigned to each MONITOR record type. Record type numbers 0-127 are reserved for class records; numbers 128-191 are reserved for DIGITAL control records; numbers 192-255 are reserved for customer control records. MON-91 MONITOR Supplemental MONITOR Information There are currently 23 record types generated by the Monitor Utility. Table MON-3 lists the MONITOR record types and their numbers, along with associated class types. (For an explanation of MONITOR class types, refer to Section A.4.1)) Table MON-3 MONITOR Record Types Type Number A.2 Record Type (in decimal) File Header 128 System Information 129 Class Type Node Transition 130 PROCESSES Class 0 component STATES Class 1 system MODES Class 2 system PAGE Class 3 system 10 Class 4 system FCP Class 5 system POOL Class 6 system LOCK Class 7 system DECNET Class 8 system RESERVED 9 system RESERVED 10 system FILE_SYSTEM_CACHE Class 1 system DISK Class 12 component RESERVED 13 component DLOCK Class 14 system SCS Class 15 component RESERVED 16 system SYSTEM Class 17 system RESERVED 18 system CLUSTER Class 19 system Conventions The following sections define the contents of each field within each record type. Record type and record size are given in decimal representation. References to “system time” indicate time values in system time format, that is, in 64-bit format. The field offset names listed are not defined within the Monitor Utility. However, DIGITAL recommends that you define and use these offset names when you work with MONITOR output records. The suggested naming convention for the field offset names appears below. Each name is of the form: MNR_CCC$X_DDDDD CCC is a record type or class mnemonic. MON-92 MONITOR Supplemental MONITOR Information X is a one-letter code indicating the size of the data item as follows: B for byte W for word L for longword Q for quadword O for octaword T for ASCII string DDDDD is the name describing the data item. In the following tables, the size of the data is also shown in parentheses, following the description of the field contents. A.3 DIGITAL Control Records There are three types of DIGITAL control records: the file header record, the system information record, and the node transition record. Each file has one header record, which contains information applicable to all classes of performance data contained in the file. It must be the first record (except for customer control records) in the file. There is one system information record per node per file. The record contains information about the VAX/VMS system being monitored and follows the header record in the file. A.3.1 File Header Record The file header record has a record type of 128 and a size of 515 bytes. Figure MON-1 illustrates the format of the file header record; Table MON-4 describes the fields in this record. MON-93 MONITOR Supplemental MONITOR Information Figure MON-1 MNR_HDRSL_FLAGS File Header Record Format Fag aie Tipe | mwA_HoRse. Tvee Flag bits Beginning Time ® byen endng Time @ tres MNR_HDRSQ_BEGINNING MNR_HDRSQ_ENOING ] MNR_HORSL. INTERVAL el Revsion Lovl 0 L INR_HDRS$O_ REVOCLSBITS 16 bytes) MINA_HDRSO.| Count MNR_HORSL_RECGT Count © A D MNR_HORST_IDENT Comments MNR_ HORST._COMMENT Comment Longh Clnssos MNR_HDRSO_ CLASSBITS Revision Levels L, MNR_HDRST_REVLEVELS (16 byis) A I MNR_HORSW_COMLEN (128 bytes) MNR_HDRSA_NODE_LIST NODE 0 16 bytes NODE 1 16 byees ] 256 byes NODETS MON-94 16 byes MONITOR Supplemental MONITOR Information Table MON-4 Descriptions of File Header Record Fields Field Symbolic Offset Contents Type MNR_HDR$B_TYPE Record type identifier (1 byte) Flags MNR_HDRS$L _FLAGS 32 flag bits; low-order bit = bit 0. All flags Beginning MNR_HDR$Q_BEGINNING reserved to DIGITAL for future use (1 longword). System time of beginning of recording (1 quadword) Ending MNR_HDR$Q _ENDING Interval MNR_HDR$L _INTERVAL System time of end of recording (1 quadword) Interval in seconds between collections; this is the value specified by the user in the recording request. It is not necessarily equal to the exact interval value obtained by subtracting two consecutive time stamps for a given class (1 longword). Revision Level 0 MNR_HDR$O_REVOCLSBITS 128-bit string representing all classes; a bit set to 1 indicates the presence in this file of a class which is at Revision Level O and whose type number corresponds to the bit number. Low-order bit = bit 0. (1 octaword). This field is provided for compatibility with Version 3.0 files. Count MNR_HDRS$L _RECCT Count of all records in the file (1 longword) D MNR_HDR$T_IDENT MONITOR Recording File Structure Level Identification (MON19044) (8 bytes) Comment MNR_HDR$T_COMMENT Recording file description supplied by the user, including trailing blanks (60 bytes) Comment Length MNR_HDR$W_COMLEN Actual length of recording file description string specified by the user (1 word) Classes MNR_HDR$O_CLASSBITS 128-bit string representing all classes; a bit set to 1 indicates the presence in this file of the ciass whose type number corresponds to the bit number. Low-order bit = bit O (1 octaword). Revision Levels MNR_HDR$T_REVLEVELS 128-byte string consisting of a 1-byte binary revision level number for each class. A class has a revision level of O initially. For each MONITOR release, if the record definition has changed, the revision level will be increased (not necessarily by 1). Node List MNR$HDR$R_NODE _LIST Names of nodes contained within the file (256 bytes) A.3.2 System Information Record The system information record has a record type of 129 and a size of 43 bytes. Figure MON-2 illustrates the format of the system information record; Table MON-5 describes the fields in this record. MON-95 MONITOR Supplemental MONITOR Information Figure MON-2 System Information Record Format Type Flags MNR_SYISW__FLAGS Time Booted (8 bytes) Max Proc Cnt CPUs MNR_SYISW__MPCPUS [Max Proc Cnt| Node Name (16 bytes) MNR_SYI$B__TYPE MNR_SYI$Q_ BOOTTIME MNR_SYISW_MAXPRCCNT MNR__SYI$T__NODENAME Bal Set Mem MNR__SYI$B__INDEX IMPW High Lim Bal Set Mem MNR__SYISL__BALSETMEM CPU Type MPW High Lim MNR_SYISL__MPWHILM Index CPU Type MNR_SYISL__CPUTYPE ZK-1984.84 Table MON-5 Descriptions of System Information Record Fields Field Symbolic Offset Contents Type MNR_SYI$B_TYPE Type identifier (1 byte) Flags MNR_SYI$SW_FLAGS 16 flag bits; low-order bit = bit 0. If bit O is set to 1, the node on which the data was collected is a member of a VAXcluster. All other flags reserved to DIGITAL for future use. (1 word) Time Booted MNR_SYI$Q_BOOTTIME System time at which system booted (1 quadword) Max Process MNR_SYISW_ MAXPROCESSCNT system parameter value (1 word) MAXPRCCNT CPUs MNR_SYI$B_MPCPUS Number of CPUs. Contains a value of 2 when the system is a VAX~-11/782 attached processor system; otherwise, contains a value of 1 (1 byte) Node Name MNR_SYI$T_NODENAME Node name of node being monitored {counted ASCII string, Balance Set MNR_SYI$L Number of process pages to which memory can be allocated Memory BALSETMEM (1 longword) MPW High MNR_SYI$L_MPWHILIM MPW_HILIMIT system parameter value (1 longword) MNR_SYI$L_CPUTYPE CPU type code. Use $PRDEF macro for code values (1 16 bytes) Limit CPU Type longword). Index MNR_SYI$B_INDEX Identifies the position of this node in several internal MONITOR data structures (1 byte). A.3.3 Node Transition Record The node transition record has a record type of 130 and a size of 2 bytes. Figure MON-3 illustrates the format of the node transition record; Table MON-6 describes the fields in this record. MON-96 MONITOR Supplemental MONITOR Information Figure MON-3 Node Transition Record Format MNR_NTR$B__INDEX Index (Byte) Type (Byte) MNR_NTRS$B__TYPE 453285 Table MON-6 Descriptions of Node Transition Record Fields Field Symbolic Offset Contents Type MNR_NTR$B_TYPE Record type identifier—indicates node removal operation (1 byte) Index MNR_NTR$B_INDEX Identifies the position of this node in several internal MONITOR data structures (1 byte) A4 Class Records The MONITOR recording file contains one class record for each requested class per collection interval, except for the PROCESSES class. (See Section A.4.2 for more information on the PROCESSES class records.) For example, if a MONITOR user requested to record five classes (excluding PROCESSES) for a duration of 100 collection intervals, the file would contain 500 class records. Class records occur in order of increasing type number within an interval. The first class record for a given interval follows the last class record for the previous interval. A.4.1 Class Type Formats There are two basic class types: system classes and component classes. A class record for a system class generally consists of counts for systemwide activities (such as page faults), whereas a class record for a component class normally contains a count for each element of a measured activity (such as 1/0 operations for each disk in the system). Specifically, a class record for a system class consists of a class header followed by a data block. A class record for a component class has a class header followed by a class prefix and one data block per element. Figure MON-4 illustrates the format for class records. MON-97 MONITOR Supplemental MONITOR Information Figure MON-4 Class Record Format SYSTEM CLASSES COMPONENT CLASSES Class Header Class Header Class Prefix Data Block Data Block 1 Data Block Data Block 2ZK-968-82 A41.1 Class Header The class header is the first part of every class record. Its format is independent of class. The class header is 13 bytes long. Figure MON-5 illustrates the format of the class header; Table MON-7 describes the fields in the class header. Figure MON-5 Class Header Format L flags MNR_CLS$B_FLAGS index [ type time recorded MNR_CLS$B_INDEX reserved MNR_CLS$Q_STAMP (8 bytes) reserved 2K-1983-84 MON-98 MNR_CLS$B_TYPE MNR_CLS$W_RESERVED MONITOR Supplemental MONITOR Information Table MON-7 Descriptions of Class Header Fields Field Symbolic Offset Contents Type MNR_CLS$B_TYPE Record type identifier (1 byte) Flags MNR_CLS$B_FLAGS Reserved MNR_CLS$W_RESERVED Reserved for DIGITAL use (1 word) Time MNR_CLS$Q_STAMP System time at which this class record was recorded. 16 flag bits; low order bit = bit 0. If bit O is set to 1, the data for this interval continues in the next record. Can be set for the PROCESSES class only. All other flags reserved by DIGITAL for future use (I byte). The time value is nondecreasing across all class records in the file. Index MNR_CLS$W_INDEX Identifies the position of this node in several internal MONITOR data structures (1 word) A4.1.2 Class Prefix (Component Classes Only) The class prefix always follows the class header for component class records. It contains data describing the number of elements (that is, processes for the PROCESSES class, disks for the DISK class, and so forth) represented by the class records for the current collection interval. Unlike system class records, which have one data block per record, component classes have one data block per element. One of the class prefix data items describes the number of elements (and therefore the number of data blocks) included in the class record. The other class prefix data item is used only for the PROCESSES class, and describes the number of processes included in the interval. The following discussion applies only to the PROCESSES class. It is possible to monitor a number of processes sufficiently large that the required number of data blocks for one collection interval will not fit into a single maximum size record. In this case, the required number of PROCESSES class records is created to fully describe the processes. All class headers in the set of PROCESSES class records for a given interval are identical, except for the setting of bit 0 in the MNR_CLS$W_FLAGS field. This bit is set to 1 for all records except the last, for which it is set to 0. The class prefixes in the set of class records vary, as described in Table MON-8. The contents of the MNR_CMP$L _ELTCT field depends on the number of data blocks contained in the record; the contents of the MNR_CMP$L_PCTINT field remains constant for each record in the set. All records in the set except the last contain as many data blocks as will fit into the maximum size record (32000 bytes). The last record in the set contains the remaining data blocks. Figure MON-6 illustrates the class prefix format; Table MON-8 describes the fields in the class prefix. The class prefix is 8 bytes long. Figure MON-6 Class Prefix Format Elements in Record MNR_CMPS$L__ELTCT Processes in Interval MNR_CMP$L__PCTINT 2K-970-82 MON-99 MONITOR Supplemental MONITOR Information Table MON-8 Descriptions of Class Prefix Fields Field Symbolic Offset Contents Elements In MNR_CMP$L_ELTCT Count of elements (data blocks) in this record (1 longword) MNR_CMPSL_PCTINT Count of processes (data blocks) for this interval (1 longword). This field is for the PROCESSES class only. Record Processes In Interval For other component classes, this longword is reserved to DIGITAL for future use. A413 Data Block The size and format of each data block and the number of blocks per record depend on the class. System classes have one data block per record. Component classes have one data block per element. The fields within each block are performance data items. The following sections describe the data items within the data block for each class. Every data item falls into one of three categories. It is either a count, a level, or an informational item. A count is a numeric quantity that increases at each succeeding interval for the life of a system boot. A level is a numeric quantity that may increase or decrease at each succeeding interval. An informational item represents data that, rather than being a unit of performance measurement (as are the first two types), is descriptive in nature, In the tables that follow, item types are identified by the letters C (count), L (level), and I (informational). The item types are shown in parentheses, following the length of the field. A.4.2 PROCESSES Class Record The PROCESSES class record contains data describing all processes in the system. The PROCESSES class record has a record type of 0; its size depends on the number of processes being monitored. The size, in bytes, is calculated by adding the size of the class header, the class prefix, and the data blocks contained in the record. This is shown by the following formula: 13 + 8 + (69 + the value of MNR_CMP$L_ELTCT) Figure MON-7 illustrates the format of the PROCESSES class record. Table MON-9 describes the fields in the data block for the PROCESSES class record. MON-100 MONITOR Supplemental MONITOR Information Figure MON-7 PROCESSES Class Record Format Class Header (13 bytes) Class Prefix (8 bytes) Internal Process ID uic Priority MNR_PROS$B__PRI MNR_PROSL_IPID Process ID State uic I MNR_PROSL_UIC MNR_PROSW__STATE Name MNR_PROST_LNAME (16 bytes) DATA Process Page Count MNR_PROSW__PPGCNT ] Global Page Count > BLOCK MNR_PROSW_GPGCNT Status Flags MNR_PROSL_STS Direct I/Os MNR_PRO$L__DIOCNT Page Faults MNR_PROSL__PAGEFLTS CPU Time MNR_PROSL_CPUTIM Buffered I/0s 1 MNR_PROSL__BIOCNT Extended Process 1D MNR_PROSL__EPID Event Flag Wait Mask MNR_PROSL_EFWM J o » H < Data Block n ~ P MON-101 MONITOR Supplemental MONITOR Information Table MON-9 Descriptions of PROCESSES Class Record Fields Field Symbolic Offset Contents Internal Process ID MNR_PROSL _IPID Internal process identifcation (longword,|) uic MNR_PROS$L _UIC User identification code (Group is high-order word; Member is low-order word) (longword,l) State MNR_PRO$W_STATE Current scheduling state code (word,|) Priority MNR_PRO$B_PRI Current software priority (complement of 31) (byte,l) Name MNR_PRO$T_LNAME Process name (counted ASCII string) (16 bytes,l) Global Page Cnt MNR_PRO$W_GPGCNT Current global page count (word,L) Process Page Cnt MNR_PRO$W_PPGCNT Current process page count (word,L) Status Flags MNR_PROSL _STS Software process status flags (PCB$V_RES bit clear implies swapped out) {longword,|) Direct 1/0s MNR_PROSL _DIOCNT Direct 1/O count (0 if swapped out) (longword,C) Page Faults MNR_PRO$L _PAGEFLTS Page fault count (0 if swapped out) (longword,C) CPU Time MNR_PROS$L _CPUTIM Accumulated CPU time, in 10ms ticks (O if swapped Buffered 1/0s MNR_PRO$L _BIOCNT Buffered 1/0 count (0 if swapped out) (longword,C) Extended Proc ID MNR_PROSL _EPID Extended process identification (longword,l) Event Fig Wt Mask MNR_PRO$L_EFWM out) (longword,C) Event flag wait mask (used for MWAITSs) (longword, 1) A.4.3 STATES Class Record The STATES class record contains data describing the number of processes in each of the scheduler states. The STATES class record has a record type of 1 and a size of 69 bytes. Figure MON-8 illustrates the format of the STATES class record. Table MON-10 describes the fields in the data block for the STATES class record. MON-102 MONITOR Supplemental MONITOR Information Figure MON-8 STATES Class Record Format Class Header (13 bytes) HEADER Collided Page Wait MNR_STASL_COLPG Misc. Resource Wait MNR_STASL__MWAIT Common Event Flag MNR_STASL_CEF Page Fault Wait MNR_STASL_PFW Local Event Flag, Inswapped MNR_STASL__LEF Local Event Flag, Outswapped - DATA BLOCK MNR_STASL_LEFO Hibernate, Inswapped MNR_STASL_HIB Hibernate, Outswapped MNR_STASL__HIBO Suspended, Inswapped MNR_STASL__SUSP Suspended, Outswapped MNR_STASL_SUSPO Free Page Wait MNR_STASL_FPG Compute State, Inswapped MNR_STASL_COM Compute State, Outswapped MNR_STASL_COMO Current MNR_STASL_CUR ZK-972-82 Table MON-10 Descriptions of STATES Class Record Fields Field Symbolic Offset Contents Collided MNR_STAS$L _COLPG Number of processes in collided page wait (longword,L) Misc Resource Wait MNR_STASL_MWAIT Number of processes in miscellaneous resource wait Common Event MNR_STASL _CEF Number of processes in common event flag wait Page Wait (longword,L) (longword,L) Flag Wait Page Fault MNR_STASL _PFW Number of processes in page fault wait (longword,L) MNR_STASL _LEF Number of processes in local event flag wait, inswapped Wait Loc Event Flag, (longword,L) Inswapped MNR_STAS$L _LEFO Number of processes in local event flag wait, outswapped (longword,L) Hibernate Inswapped MNR_STAS$L_HIB Number of processes in hibernate wait, inswapped Hibernate Outswapped MNR_STA$L _HIBO Loc Event Flag, . Outswapped (longword,L) Number of processes in hibernate wait, outswapped (longword,L) MON-103 MONITOR Supplemental MONITOR Information Table MON-10 (Cont.) Descriptions of STATES Class Record Fields Field Symbolic Offset Contents Suspended MNR_STAS$L _SUSP Number of processes in suspended wait, inswapped Inswapped (longword,L) Suspended MNR_STAS$L_SUSPO Free Page MNR_STASL_FPG Number of processes in free wait (longword,L) Compute State MNR_STAS$L_COM Number of processes in compute state, inswapped Compute State Outswapped MNR_STAS$L_COMO Number of processes in compute state, outswapped (longword,L) Current MNR_STASL_CUR Number of current processes (longword,L) Outswapped Number of processes in suspended wait, outswapped (longword,L) Wait Inswapped A.4.4 (longword,L) MODES Class Record The MODES class record contains data describing time spent in each of the processor modes. The MODES class record has a record type of 2. Its size is 41 bytes when monitoring a single processor, and 69 bytes when monitoring two processors (VAX-11/782). Figure MON-9 illustrates the format of the MODES class record. Table MON-11 describes the fields in the data block for the MODES class record. The first seven fields refer to the primary processor. The last seven fields refer to the attached processor; they are included in the MODES record only if the MNR_SYI$B_MPCPUS field of the System Information Record contains a value of 2, indicating that two processors are being monitored. MON-104 MONITOR Supplemental MONITOR Information Figure MON-9 HEADER MODES Class Record Format C'(a‘s;:yf::f' L | Interrupt Stack MNR_MODSL_INTER Kenel MNR_MODSL_KERNEL Executive Supervisor User Compatibility DATA idle o Interrupt Stack MNR_MODSL_EXEC MNR_MODSL_SUPER MNR_MODSL_USER MNR_MODSL_COMPAT MNR_MODSL_IDLE MNR_MODSL_INTER_S Kernel MNR_MODSL_KERNEL_S Executive MNR_MODSL_EXEC_S Supervisor MNR_MODSL_SUPER_S User MNR_MODSL_USER_S Compatibility dle RLAGTL MNR_MODSL_COMPAT_S MNR_MODSL_IDLE_S wers Table MON-11 Field Descriptions of MODES Class Record Fields Symbolic Offset Contents Fields Referred to Primary Processor MNR_MODSL _INTER Count of clock ticks (10-millisecond units) spent on Kernel MNR_MODS$L _KERNEL Count of clock ticks since system boot spent in kernel Executive MNR_MODS$L _EXEC Count of clock ticks since system boot spent in Supervisor MNR_MODS$L _SUPER Count of clock ticks since system boot spent in supervisor mode (longword,C) User MNR_MODS$L _USER Count of clock ticks since system boot spent in Interrupt Stack interrupt stack since system was booted (longword,C) mode, excluding interrupt stack time (longword,C) executive mode (longword,C) user mode, excluding compatibility-mode time (longword,C) Compatibility MNR_MOD$L _COMPAT Count of clock ticks since system boot spent in compatibility mode (longword,C) MON-105 MONITOR Supplemental MONITOR Information Table MON-11 (Cont.) Descriptions of MODES Class Record Fields Field Symbolic Offset Idle MNR_MODS$L _IDLE Contents Count of clock ticks since system boot spent executing the NULL process (longword,C) Fields Referred to Attached Processor Interrupt Stack MNR_MODS$L _INTER_S Count of clock ticks (10-millisecond units) spend on interrupt stack since system was booted (longword,C) Kernel MNR_MODS$L _KERNEL _S Count of clock ticks since system boot spent in kernel mode, excluding interrupt stack time (longword,C) Executive MNR_MODS$L _EXEC_S Count of clock ticks since system boot spent in executive mode (longword,C) Supervisor MNR_MODS$L _SUPER_S Count of clock ticks since system boot spent in supervisor mode (longword,C) User MNR_MOD$L _USER_S Count of cluck ticks since system boot spent in user mode, excluding compatibility-mode time ({longword,C) Compatibility MNR_MOD$L_COMPAT_S Count of clock ticks since system boot spent in compatibility mode (longword,C) Idle MNR_MODS$L _IDLE_S Count of clock ticks since system boot spent executing the NULL process (longword,C) A.4.5 PAGE Class Record The PAGE class record contains data describing the operation of the page management subsystem. The PAGE class record has a record type of 3 and a size of 65 bytes. Figure MON-10 illustrates the format of the PAGE class record. Table MON-12 describes the fields in the data block for the PAGE class record. MON-106 MONITOR Supplemental MONITOR Information Figure MON-10 PAGE Class Record Format HEADER c'fis::;::f’ o EEmeaRL] = . Page Faults MNR_PAGSL_FAULTS Read 1/0s MNR_PAGSL_PREADIO Reads Writes MNR_PAGSL_PREADS MNR_PAGSL_PWRITES Write 1/0s MNR_PAGSL_PWRITIO Free Page List Faults MNR_PAGSL_FREFLTS DATA Modified Page List Faults MNR_PAGSL_MFYFLTS BLOCK Demand-Zero Faults MNR_PAGSL_DZROFLTS Global Valid Faulls MNR_PAGSL_GVALID Wite in Progress Faults MNR_PAGSL_WRTINPROG System Faults MNR_PAGSL_SYSFAULTS Free Page Count MNR_PAGSL_FREECNT Modified Page Count MNR_PAGSL_MFYCNT 2K974-82 Table MON-12 Descriptions of PAGE Class Record Fields Contents Field Symbolic Offset Page Faults MNR_PAGSL _FAULTS Count of page faults for all working set (longword,C) Reads MNR_PAG$L _PREADS Count of pages read from disk as a result of page faults (longword,C) Read 1/0s MNR_PAGS$L _PREADIO Count of read 1/Os as a result of operations from disk page faults (longword,C) Writes MNR_PAGS$L _PWRITES Count of pages written to the page file (longword,C) Write 1/Os MNR_PAGSL _PWRITIO Count of write 1/0 operations to the page file Free Page MNR_PAGSL _FREFLTS Count of pages read from the free list as a result of MNR_PAG$L _MFYFLTS Count of pages read from the modified list as a MNR_PAGS$L _DZROFLTS Count of zero-filled pages allocated as a result of (longword,C) List Faults Modified Page List Faults Demand-zero Faults result of page faults (longword,C) faults (longword,C) Faults Global Valid page faults (longword,C) MNR_PAGSL _GVALID Count of page faults for which the reference page was found to be valid in the system global page tables (longword,C) MON-107 MONITOR Supplemental MONITOR Information Table MON-12 (Cont.) Descriptions of PAGE Class Record Fields Field Symbolic Offset Contents Write in Progress MNR_PAGSL _ Count of pages read that were in the process System Faults MNR_PAGSL . Count of page faults for which the referenced page SYSFAULTS is in system space (longword,C) Free Page Count MNR_PAGS$L _FREECNT Number of pages currently on free page list Modified Page MNR_PAG$L _MFYCNT Faults WRTINPROG of being written back to disk when faulted (longword,C) (longword,L) Count A.4.6 Number of pages currently on modified page list (longword,L) 1/0 Class Record The 1/0 class record contains data describing the operation of the I/O subsystem. The /O class record has a record type of 4 and a size of 69 bytes. Figure MON-11 illustrates the format of the I/O class record. Table MON-13 describes the fields in the data block for the I/O class record. Figure MON-11 HEADER 1/0 Class Record Format Class Header (13 bytes) Direct I/Os MNR_IOS$L_DIRIO Buffered I/0s MNR_IOS$L__BUFIO Mailbox Writes MNR_IOS$L_MBWRITES Window Turns MNR_IOSL__FCPTURN Logical Name Translations MNR_IOSL_LOGNAM Files Opened MNR_IO$L__OPENS DATA Page Faults MNR_IOSL__FAULTS BLOCK Page Reads MNR_IOSL__PREADS Page Read I/Os MNR_IO$L__PREADIO Page Writes MNR_IO$L__PWRITES Page Write I/0s MNR_IOSL_PWRITIO Inswaps MNR_IOS$L__ISWPCNT Free Page Count MNR_IO$L_FREECNT Modified Page Count MN R_lOsl_MFVC".lT 2K-975.82 MON-108 MONITOR Supplemental MONITOR Information Table MON-13 Descriptions of 1/0 Class Record Fields Field Symbolic Offset Contents Direct 1/0s MNR_IO$L _DIRIO Count of direct 1/O operations {longword,C) Buffered 1/Os MNR_IO$L _BUFIO Count of buffered 1/O operations (longword,C) Mailbox Writes MNR_IO$L_MBWRITES Count of write-to-mailbox requests (longword,C) Window Turns MNR_IO$L _FCPTURN Count of file-map window misses (longword,C) Logical Name MNR_IO$L _LOGNAM Count of logical name translations (longword,C) Files Opened MNR_IO$L _OPENS Count of files opened {longword,C) Page Faults MNR_IO$L_FAULTS Count of page faults for all working sets (longword,C) Page Reads MNR_IO$L _PREADS Translations Count of pages read from disk as a result of page faults (longword,C) Count of read |/O operations from disk as a result of Page Read 1/0s MNR_IO$L _PREADIO Page Writes MNR_IO$L _PWRITES Count of pages written to the page file (longword,C) Page Write 1/0s MNR_IO$L_PWRITIO Count of write 1/O operations to the page file page faults (longword,C) (longword,C) Inswaps MNR_IO$L _ISWPCNT Free Page Count MNR_IO$L _FREECNT Count of working sets read into memory from the swap file (longword,C) Number of pages currently on free page list (longword,L) Modified Page MNR_IO$L_MFYCNT A4.7 Number of pages currently on modified page list (longword,L) Count FCP Class Record The FCP class record contains data describing the operation of the file system ACPs. The FCP class record has a record type of 5 and a size of 61 bytes. Figure MON-12 illustrates the format of the FCP class record. Table MON-14 describes the fields in the data block for the FCP class record. MON-109 MONITOR Supplemental MONITOR Information Figure MON-12 - FCP Class Record Format HEADER f C'(‘:s; :;:;e' L | ( FCP Calls Disk Allocations RECORD < DATA BLOCK L L MNR_FCPSL_FCPCALLS MNR_FCPSL_ALLOC New Files MNR_FCPS$L_FCPCREATE Read I/Os MNR_FCPSL_FCPREAD Wiite 1/0s MNR_FCPSL_FCPWRITE Volume Lock Waits MNR__FCP$L_VOLWAIT CPU Time MNR__FCP$L_FCPCPU Page Faults MNR_FCP$L_FCPFAULT Window Turns MNR_FCPSL_FCPTURN Access MNR_FCPSL_ACCESS Files Opened MNR_FCP$L_OPENS Erase 1/0 Operations MNR_FCP$L__ERASE 2K-976-82 Table MON-14 Descriptions of FCP Class Record Fields Field Symbolic Offset FCP Calls MNR_FCP$L _FCPCALLS Contents Count of QIO requests received by the file system (longword,C) Disk Allocations MNR_FCP$L_ALLOC Count of QIO requests that caused allocation of disk space (longword,C) New Files MNR_FCP$L _ Count of new files created (longword,C) FCPCREATE Read 1/0s MNR_FCP$L_FCPREAD Count of read 1/O operations from the disk by the file system (longword,C) Write 1/0s MNR_FCP$L _FCPWRITE Count of write /O operations to disk by the file system (longword,C) Volume Lock Waits MNR_FCP$L_VOLWAIT Number of times a wait state was entered by the XQP due to volume lock contention (longword,C) CPU Time MNR_FCP$L_FCPCPU FCP Page Faults MNR_FCP$L _FCPFAULT Count of page faults for the file system (longword,C) Window Turns MNR_FCP$L _FCPTURN Count of file-map window misses (longword,C) Count of clock ticks (10-millisecond units) of CPU time used by the file system (longword,C) MON-110 MONITOR Supplemental MONITOR Information Table MON-14 (Cont.) Descriptions of FCP Class Record Fields Field Symbolic Offset Contents Access MNR_FCP$L __ACCESS Count of filename lookup operations in file Files Opened Erase 1/0 MNR_FCP$L_OPENS MNR_FCP$L _ERASE Count of files opened (longword,C) Count of erase 1/0 operations issued (longword,C) directories (longword,C) Operations A.4.8 POOL Class Record The POOL class record contains data describing space allocation in the nonpaged dynamic pool. The POOL class record has a record type of 6 and a size of 61 bytes. Figure MON-13 illustrates the format of the POOL class record. Table MON-15 describes the fields in the data block for the POOL class record. Figure MON-13 POOL Class Record Format Class Header (13 bytes) E Small Request Packets Free MNR_POO$L__SRPCNT Smali Request Packets In Use MNR_POOS$L__SRPINUSE Intermediate Request Packets Free MNR_POOSL__IRPCNT Intermediate Request Packets In Use MNR_POOSL__IRPINUSE Large Request Packets Free MNR_POOS$L__LRPCNT Large Request Packets In Use MNR_POOS$L__LRPINUSE Unused Bytes MNR_POOS$L__HOLESUM Bytes In Use MNR_POOS$L__DYNINUSE Unused Contiguous Space MNR_POO$L_HOLECNT Largest Block MNR_POOS$L__BIGHOLE Smallest Block MNR_POO$L_SMALLHOLE Blocks Less Than Or Equal to 32 MNR_POOS$L_SMALLCNT 2K-1988-84 MON-111 MONITOR Supplemental MONITOR Information Table MON-15 Descriptions of POOL Class Record Fields Field Symbolic Offset Contents Small Request Packets MNR_POOSL _SRPCNT Number of preallocated small request packets Available left in lookaside list (longword,L) MNR_POOSL _SRPINUSE Small Request Packets In Use Number of preallocated small request packets in use (longword,L) MNR_POOS$L _IRPCNT Intermediate Request Packets Available Number of preallocated intermediate request packets left in lookaside list (longword,L) MNR_POOSL _IRPINUSE Intermediate Request Packets In Use Number of preallocated intermediate request packets in use (longword,L) Large Request Packets MNR_POO$L _LRPCNT Number of preallocated large request packets MNR_POOS$L _LRPINUSE Number of preallocated large request packets in use (longword,L) MNR_POOS$L _HOLESUM Total unused bytes in the dynamically allocated portion of nonpaged pool (longword,L) MNR_POOSL _DYNINUSE Number of bytes in the dynamically allocated Available left in lookaside list (longword,L) Large Request Packets in Use Unused Bytes In Dynamic Nonpaged Pool Bytes Of Dynamic Nonpaged Pool In Use portion of nonpaged pool (longword,L) Unused Contiguous Space MNR_POOSL _HOLECNT Largest Block MNR_POOSL _BIGHOLE Number of unused blocks of contiguous space in the dynamically allocated portion of nonpaged pool (longword,L) Size in bytes of the largest block of unused space in the dynamically allocated portion of nonpaged pool (longword,L) MNR_POOSL Smallest Block SMALLHOLE Size in bytes of the smallest block of unused space in the dynamically allocated portion of nonpaged pool (longword,L) Blocks Less Than Or MNR_POOSL _SMALLCNT Equal to 32 Number of blocks less than or equal to 32 bytes in size in the dynamically allocated portion of nonpaged pool (longword,L) A.4.9 LOCK Class Record The LOCK class record contains data describing the operation of the lock management subsystem. The LOCK class record has a record type of 7 and a size of 53 bytes. Figure MON-14 illustrates the format of the LOCK class record. Table MON-16 describes the fields in the data block for the LOCK class record. MON-112 MONITOR Supplemental MONITOR Information Figure MON-14 LOCK Class Record Format HEADER Oe L | New ENQs MNR_LCKSL_ENQNEW Converted ENQs MNR_LCKSL_ENQCVT DEQs MNR_LCKSL_DEQ DATA Blocking ASTs MNR_LCKSL_BLKAST BLOCK ENQ Waits MNR_LCKSL_ENQWAIT ENQs Not Queued MNR_LCKSL_ENQNOTQD Deadock Searches MNR_LCKSL_DLCKSRCH Deadlocks Found MNR_LCKSL_DLCKFND Current Locks MNR_LCKSL_NUMLOCKS Current Resources MNR_LCKSL_NUMRES zk-978-82 Table MON-16 Descriptions of LOCK Record Fields Field Symbolic Offset Contents New ENQs MNR_LCK$L _ENQNEW Count of new ENQ (lock) requests (longword,C) Converted ENQs MNR_LCK$L _ENQCVT Count of converted ENQ (lock) requests (longword,C) DEQs MNR_LCK$L _DEQ Count of DEQ (unlock) requests (longword,C) Blocking ASTs MNR_LCK$L _BLKAST Count of blocking ASTs queued (longword,C) ENQ Waits MNR_LCK$L _ENQWAIT Count of times a lock could not be granted immediately and waited (longword,C) ENQs Not Queued MNR_LCK$L _ENQNOTQD Count of times a lock could not be granted immediately and got an error status instead of waiting {longword,C) Deadlock Searches MNR_LCK$L _DLCKSRCH Count of times that a deadlock search was performed Deadlocks Found MNR_LCK$L _DLCKFND Count of times that a deadlock was found (longword,C) Current Locks MNR_LCK$L _NUMLOCKS Number of locks currently in the system (longword,L) Current Resources MNR_LCK$L _NUMRES Number of resources currently in the system (longword,C) (longword,L) A.4.10 DECNET Class Record The DECNET class record contains data describing the operation of the DECnet-VAX subsystem. The DECNET class record has a record type of 8 and a size of 37 bytes. Figure MON-15 illustrates the format of the DECNET class record. Table MON-17 describes the fields in the data block for the DECNET class record. MON-113 MONITOR Supplemental MONITOR Information Figure MON-15 HEADER DECNET Class Record Format Oy oo L DATA BLOCK Arriving Local Packets MNR_NETSL_ARRLOCPK Departing Local Packets MNR_NET$L__DEPLOCPK Arriving Transit Packets MNR_NETSL_ARRTRAPK Transit Packets Lost MNR_NETSL_TRCNGLOS Receiver Buffer Failures MNR_NET$L__RCVBUFFL Large Request Packets MNR_NETSL__LRPCNT 2K-979.82 Table MON-17 Descriptions of DECNET Class Record Fields Field Symbolic Offset Contents Arriving MNR_NET$L_ARRLOCPK Count of arriving local packets (longword,C) MNR_NET$L_DEPLOCPK Count of departing local packets (longword,C) MNR_NET$L_ARRTRAPK Count of arriving transit packets (longword,C) MNR_NET$L_TRCNGLOS Count of packets lost because of transit congestion MNR_NETSL_RCVBUFFL Count of receiver buffer failures (longword,C) Local Packets Departing Local Packets Avrriving Transit Packets Transit Packets Lost Receiver Buffer (longword,C) Failures Large Request Packets MNR_NET$L _LRPCNT Number of preallocated large request packets left in nonpaged pool (longword,L) A.4.11 FILE_SYSTEM_CACHE Class Record The FILE_SYSTEM_CACHE class record contains data describing the operation of the caches for the file system ACPs and XQPs . The FILE_ SYSTEM _CACHE class record has a record type of 11 and a size of 69 bytes. Figure MON-16 illustrates the format of the FILE_SYSTEM_CACHE class record. Table MON-18 describes the fields in the data block for the FILE SYSTEM._CACHE class record. MON-114 MONITOR Supplemental MONITOR Information Figure MON-16 FILE_SYSTEM_CACHE Class Record Format Class Header (13 bytes) e Sl Directory FCB Cache Hits MNR_FILSL_DIRFCB__HIT Directory FCB Cache Attempts Directory Data Cache Hits MNR_FIL$L__DIRFCB__TRIES MNR_FILSL_DIRDATA_HIT Directory Data Cache Attempts MNR_FIL$L__DIRDATA_TRIES MNR_FILSL_FILHDR_HIT File Header Cache Hits File Header Cache Attempts MNR_FILSL_FILHDR__TRIES MNR_FILSL__FIDHIT File ID Cache Hits File ID Cache Attempts MNR_FILSL__FIDZ TRIES Extent Cache Hits MNR_FILSL_EXTHIT MNR_FILSL_EXT_TRIES Extent Cache Attempts Quota Cache Hits MNR_FILSL_QUOHIT MNR_FILSL_QUO_TRIES Quota Cache Attempts Storage Bitmap Cache Hits Storage Bitmap Cache Attempts, MNR_FIL$L_STORAGMAP_HIT MNR_FIL$L_STORAGMAP_TRIES 2K-1085-84 Table MON-18 Descriptions of FILE_SYSTEM_CACHE Class Record Fields Field Symbolic Offset Contents Directory FCB MNR_FILSL _DIRFCB_HIT Count of hits on Directory FCB cache Directory FCB MNR_FILSL _DIRFCB_TRIES Count of attempts on Directory FCB cache Cache Hits (longword,C) (longword,C) Cache Attempts Directory Data Cache Hits MNR_FIL$L _DIRDATA_HIT Count of hits on Directory Data cache Directory Data Cache Attempts MNR_FIL$L _DIRDATA_TRIES Count of attempts on Directory Data cache File Header Cache Hits MNR_FILSL _FILHDR_HIT File Header MNR_FIL$L _FILHDR_TRIES (longword,C) (longword,C) Count of hits on File Header cache (longword,C) Count of attempts on File Header cache {longword,C) Cache Attempts MNR_FILSL _FIDHIT Count of hits on File ID cache (longword,C) File ID Cache Attempts MNR_FIL$L _FID_TRIES Count of attemps on File ID cache (longword,C) Extent Cache MNR_FILSL _EXTHIT Count of hits on Extent cache (longword,C) File ID Cache Hits Hits MON-115 MONITOR Supplemental MONITOR Information Table MON-18 (Cont.) Descriptions of FILE_SYSTEM_CACHE Class Record Fields Field Symbolic Offset Contents Extent Cache MNR_FILSL _EXT_TRIES Count of attempts on Extent cache Quota Cache MNR_FILSL _QUOHIT Count of hits on Quota Cache (longword,C) Quota Cache MNR_FILSL _QUO_TRIES Count of attempts on Quota cache Storage Bitmap MNR_FILSL _STORAGMAP_HIT Count of hits on Storage Bitmap cache Attempts (longword,C) Hits Attempts Cache Hits Storage Bitmap (longword,C) (longword,C) MNR_FILSL _STORAGMAP_TRIES Count of attempts on Storage Bitmap cache (longword,C) Cache Attempts A.4.12 DISK Class Record The DISK class record contains data describing all disk devices in the system. The DISK class record has a record type of 12; its size depends on the number of disks being monitored. The size, in bytes, is calculated by adding the size of the class header, the class prefix, and the data blocks contained in the record: 13 + 8 + (36 * the value of MNR_CMP$L_ELTCT) Figure MON-17 illustrates the format of the DISK class record. Table MON-19 describes the fields in the data block for the DISK class record. MON-116 MONITOR Supplemental MONITOR Information Figure MON-17 DISK Class Record Format Class Header (13 bytes) Class Prefix (8 bytes) Allocation MNR_DSK$T__ALLOCLS Class MNR_DSK$T__CTRLR Controller MNR_DSK$W__UNITNO Unit Number MNR_DSK$T_NODENAME Volume Name MNR_DSK$T_VOLNAME Operations MNR_DSKSL_OPCNT Queue Length MNR_DSK$L__IOQUELN Data Block n n B3 3 1€ N\ Node Name 2K-1987-84 MON-117 MONITOR Supplemental MONITOR Information Table MON-19 Descriptions of DISK Class Record Fields Field Symbolic Offset Contents Allocation Class MNR_DSK$B_ALLOCLS Allocation class number (byte,l) Controller MNR_DSK$T_CTRLR Name of device controller (counted ASCII string) (4 bytes,l) Unit Number MNR_DSK$W_UNITNO Unit number (word,|) Node Name MNR_DSKS$T_ Name of cluster node where device resides (counted NODENAME ASCII string) (8 bytes,|) Volume Name MNR_DSK$T_VOLNAME Volume name of disk (ASCII) (12 bytes,|) Operations MNR_DSK$L_OPCNT Count of 1/0 operations (longword,C) Queue Length MNR_DSK$L _IOQUELN Sum of 1/0 request queue samples (longword,C) A.4.13 DLOCK Class Record The DLOCK class record contains data describing the operation of the Distributed Lock Management Facility. The DLOCK class record has a record type of 14 and a size of 73 bytes. Figure MON-18 illustrates the format of the DLOCK class record. Table MON-20 describes the fields in the data block for the DLOCK class record. MON-118 MONITOR Supplemental MONITOR Information Figure MON-18 HEADER DLOCK Class Record Format Class Header (15 bytes) MNR_DLO$L_ENQNEWLOC New Locks — Local DATA BLOCK New Locks — Incoming MNR_DLOSL_ENQNEWIN New Locks — Outgoing MNR_DLOSL_ENQNEWOUT Lock Conversions — Local MNR_DLO$L_ENQCVTLOC Lock Conversions — Incoming MNR_DLOSL_ENQCVTIN Lock Conversions — Outgoing MNR_DLOSL_ENQCVTOUT Unlocks — Local MNR_DLOS$L__DEQLOC Unlocks — Incoming MNR_DLOSL__DEQIN Unlocks — Outgoing MNR_DLO$L__DEQOUT Blocking ASTs — Local MNR_DLO$L__BLKLOC Blocking ASTs — Incoming MNR_DLOS$L__BLKIN Blocking ASTs — Outgoing MNR_DLOSL__BLKOUT Functions Incoming MNR_DLOSL__DIRIN Functions Outgoing MNR_DLO$L_DIROUT Deadlock Message Rate MNR_DLOS$L__DLCKMSG 2K-1349-83 Table MON-20 Descriptions of DLOCK Class Record Fields Field Symbolic Offset Contents New Locks MNR_DLOSL _ Count of new lock requests which originate —Local ENQNEWLOC and are performed on this system (local) ({longword,C) New Locks MNR_DLO$L _ENQNEWIN —Incoming Count of new lock requests originating on another system and performed on this system (incoming) (longword,C) New Locks —~Outgoing MNR_DLOSL ENQNEWOUT Count of new lock requests originating on this system and performed on another system Lock Conversions MNR_DLOSL _ENQCVTLOC Count of lock conversion requests (local) (outgoing) (longword,C) —Local Lock Conversions —Incoming (longword,C) MNR_DLOSL _ENQCVTIN Count of lock conversion requests (incoming) (longword,C) MON-119 MONITOR Supplemental MONITOR Information Table MON-20 (Cont.) Descriptions of DLOCK Class Record Fields Field Symbolic Offset Contents Lock Conversions MNR_DLOSL Count of lock conversion requests (outgoing) —Outgoing ENQCVTOUT (longword,C) Unlocks—Local MNR_DLOS$L_DEQLOC Count of unlock requests (local) (longword,C) Unlocks—Incoming MNR_DLOSL _DEQIN Count of unlock requests (incoming) (longword,C) Unlocks—Outgoing MNR_DLO$L _DEQOUT Count of unlock requests (outgoing) (longword,C) Blocking ASTs MNR_DLOSL _BLKLOC Count of lock manager blocking ASTs (local) Blocking ASTs MNR_DLOSL _BLKIN Count of lock manager blocking ASTs —Local —Incoming (longword,C) (incoming) (longword,C) Blocking ASTs MNR_DLOS$L _BLKOUT Count of lock manager blocking ASTs (outgoing) Directory Functions MNR_DLOS$L _DIRIN Count of directory functions (incoming) —Outgoing —Incoming (longword,C) (longword,C) Directory Functions MNR_DLOSL _DIROUT —Outgoing Count of directory functions (outgoing) (longword,C) MNR_DLOS$L _DLCKMSG Deadlock Message Rate Count of incoming and outgoing lock manager messages required for deadlock detection (longword,C) A.4.14 SCS Class Record The SCS class record contains data describing SCS (System Communication Services) activity for all SCS connections in the system, on a per-node basis. The SCS class record has a record type of 15; its size depends on the number of nodes being monitored. The size, in bytes, is calculated by adding the size of the class header, the class prefix, and the data blocks contained in the record: 13 + 8 + (66 *+ the value of MNR_CMP$L_ELTCT) Figure MON-19 illustrates the format of the SCS class record. Table MON-21 describes the fields in the data block for the SCS class record. MON-120 MONITOR Supplemental MONITOR Information Figure MON-19 HEADER PREFIX SCS Class Record Format Class Header (13 bytes) Class Prefix (8 bytes) [pasioe| MNR_SCS$T_NODENAME NODE NAME Datagrams Sent MNR_SCS$L_DGSENT Datagrams Received MNR_SCS$L_DGRCVD MNR__SCS$L_DGDISCARD Datagrams Discarded DATA BLOCKS Sequenced Messages Sent MNR_SCS$L__MSGSENT Sequenced Messages Received MNR_SCS$L_MSGRCVD Block Transfer Send-datas MNR_SCS$L__SNDATS Kilobytes Sent by Send-datas MNR_SCS$L__KBYTSENT Block Transfer Request-datas MNR_SCS$L_REQDATS Kilobytes Received by Request-datas MNR_SCS$L_KBYTREQD Block Transfer Kilobytes Mapped MNR_SCS$L_KBYTMAPD Connections Queued for Send Credit MNR_SCS$L_QCRCNT Connections Queued for Buffer Descriptor MNR_SCS$L_QBDTCNT DATA BLOCK n ) DATA BLOCK ~ a 2K-1670-84 MON-121 MONITOR Supplemental MONITOR Information Table MON-21 Descriptions of SCS Class Record Fields Field Symbolic Offset Contents Node Name MNR_SCS$T_ Name of remote cluster node (counted ASCII NODENAME string) (8 bytes,l) Datagrams Sent MNR_SCS$L _DGSENT Count of datagrams sent to the remote node Datagrams Received MNR_SCS$L_DGRCVD (longword,C) Count of datagrams received from the remote node (longword,C) MNR_SCSS$L _ Count of datagrams discarded by the Ci port DGDISCARD driver (longword,C) Sequenced Msgs Sent MNR_SCS$L_MSGSENT Count of sequenced messages sent to the Seqenced Msgs Received MNR_SCS$L_MSGRCVD Datagrams Discarded remode node (longword,C) Count of sequenced messages received from the remote node (longword,C) Block Transfer Send-datas MNR_SCS$L_SNDATS Count of block transfer send-datas initiated on the local node, targeted for the remote node (longword,C) Kilobytes Sent MNR_SCS$L _KBYTSENT by Send-datas Block Transfer Count of kilobytes sent as a result of senddatas (longword,C) MNR_SCS$L _REQDATS Request-datas Count of block transfer request-datas initiated on the local node, targeted for the remote node (longword,C) Kilobytes Received MNR_SCS$L _KBYTREQD Count of kilobytes received as a result of request-datas (longword,C) by Request-datas Block Transfer MNR_SCS$L Count of kilobytes mapped for block transfers Kilobytes Mapped KBYTMAPD (longword,C) Connections Queued For Send Credit MNR_SCS$L _QCRCNT Count of times connections are queued for send credits (longword,C) Connections Queued For Buffer Descriptor MNR_SCS$L_QBDTCNT Count of times connections are queued for buffer descriptors (longword,C) A.4.15 SYSTEM Class Record The SYSTEM class record contains data describing the overall operation of the three major system components (CPU, memory, I/O). The SYSTEM class record has a record type of 17 and a size of 49 bytes. Please note that when the SYSTEM class is recorded, the PROCESSES, STATES, and MODES classes are also recorded, even if not explicitly requested. Figure MON-20 illustrates the format of the SYSTEM class record. Table MON-22 describes the fields in the data block for the SYSTEM class record. MON-122 MONITOR Supplemental MONITOR Information Figure MON-20 SYSTEM Class Record Format Class Header (13 bytes) CPU Busy MNR_SYSS$L__BUSY Other States MNR_SYS$L__OTHSTAT Process Count MNR_SYS$L__PROCS Page Faults MNR_SYS$L__FAULTS Read 1/0s MNR_SYS$L__PREADIO Free Page Count MNR_SYS$L__FREECNT Modified Page Count MNR_SYS$L__MFYCNT Direct 1/0s MNR_SYS$L_DIRIO Buffered 1/0s MNR_SYS$L__BUFIO 2K-1986-84 Table MON-22 Descriptions of SYSTEM Class Record Fields Field Symbolic Offset Contents CPU Busy MNR_SYSS$L_BUSY Count of clock ticks (10-millisecond units) spent in all CPU modes since system was booted (longword,C) Other States MNR_SYS$L Number of processes in states other than LEF, LEFO, HIB, HIBO, COM, COMO, PFW, and MWAIT OTHSTAT (longword,L) Process Count MNR_SYS$L_PROCS Number of processes in system (longword,L) Page Faults MNR_SYS$L_FAULTS Count of page faults for all working sets (longword,C) Read 1/0s MNR_SYS$L_PREADIO Count of read |/Os resulting from disk page faults (longword,C) Free Page Count MNR_SYSS$L_FREECNT Number of pages currently on free page list Modified Page Count MNR_SYS$L_MFYCNT Number of pages currently on modified page list (longword,L) (longword,L) Direct 1/0s MNR_SYS$L _DIRIO Count of direct 1/0 operations (longword,C) Buffered 1/Os MNR_SYS$L _BUFIO Count of buffered 1/0 operations (longword,C) MON-123 MONITOR Supplemental MONITOR Information A.4.16 CLUSTER Class Record The CLUSTER class record contains data describing clusterwide CPU, memory, and locking activity. The CLUSTER class record has a record type of 19 and a size of 65 bytes. Please note that when the CLUSTER class is recorded, the DISK and MODES classes are also recorded, even if not explicitly requested. Figure MON-21 illustrates the format of the CLUSTER class record. Table MON-23 describes the fields in the data block for the CLUSTER class record. Figure MON-21 HEADER CLUSTER Class Record Format Class Header (13 bytes) L CPU Busy Free List Size Reserved Total Locks New ENQ Local MNR__CLUSL_CPU__BUSY MNR_ CLUSL__FRLIST MNR__CLUSL__RESERVED MNR_CLUSL_TOTAL_LOCKS MNR_CLUSL_ENQNEWLOC New ENQ Incoming MNR__CLUSL_ENQNEWIN New ENQ Outgoing MNR_CLUSL_ENQNEWOUT ENQ Conversions Local MNR__CLUSL_ENQCVTLOC ENQ Conversions Incoming MNR__CLUSL_ENQCVTIN ENQ Conversions Outgoing MNR_CLUSL_ENQCVTOUT DEQ Local MNR_CLUSL__DEQLOC DEQ Incoming MNR__CLUSL_DEQIN DEQ Outgoing MNR__CLUSL_DEQOUT 2453185 MON-124 MONITOR Supplemental MONITOR Information Table MON-23 Descriptions of CLUSTER Class Record Fields Field Symbolic Offset Contents CPU Busy MNR_CLUS$L _CPU_BUSY Count of clock ticks (10-millisecond Free List Size MNR_CLUSL _FRLIST units) spent in all CPU modes since system was booted (longword,C) Number of pages currently on the free list (longword,L} Reserved MNR_CLU$L _RESERVED Reserved to DIGITAL Total Locks MNR_CLUSL_TOTAL _LOCKS Total of all incoming, outgoing, and local ENQs, DEQs, and conversions (longword,C) New Enq Local MNR_CLUSL _ENQNEWLOC Count of new lock requests which originate and are performed on the system (local) (longword,C) New Enq Incoming MNR_CLUSL _ENQNEWIN Count of new lock requests which originate on other systems and are performed on this system (incoming) (longword,C) New Eng Outgoing MNR_CLUSL _ENQNEWOUT Count of new lock requests which originate on this system and are performed on other systems (outgoing) (longword,C) Enq Conversions Local MNR_CLUSL _ENQCVTLOC Count of lock conversion requests (local) (longword,C) Enqg Conversions Incoming MNR_CLU$L _ENQCVTIN Count of lock conversion requests (incoming) {longword,C) Eng Conversions Outgoing MNR_CLU$L _ENQCVTOUT Count of lock conversion requests (outgoing) (longword,C) Deq Local MNR_CLU$L _DEQLOC Count of unlock requests (local) Deq Incoming MNR_CLU$L _DEQIN Count of unlock requests (incoming) (longword,C) Deq Outgoing MNR_CLU$L _DEQOUT Count of unlock requests (outgoing) ({longword,C) (longword,C) MON-125 O Index Example concurrent display and recording ®* MON-85 live display monitoring® MON-84 /BEGINNING qualifier *MON-18 /BY_NODE qualifier * MON-19 live recording ® MON-84 playback * MON-86 remote playback ® MON-87 rerecording ® MON-87 C Class-name qualifier * MON-4 VAXcluster multifile summary MON-88 EXECUTE command ® MON-34 EXIT command ® MON-35 Exiting from MONITOR ® MON-2 /ALL*MON-38 /AVERAGE ®* MON-38 /CPU*MON-62 F /CURRENT ® MON-38 /ITEM®MON-47, MON-72 /MAXIMUM ¢ MON-38 /MINIMUM ¢ MON-38 /PERCENT ® MON-47 File primitive statistics® MON-52 File space ® MON-11 File system cache (ACP) SYSGEN parameters® MON-57 /TOPBIO ® MON-69 File system cache statistics * MON-54 /TOPCPU ® MON-69 /FLUSH_INTERVAL qualifiereMON-23 /TOPDIO ® MON-69 /TOPFAULT ® MON-69 Class type component classes®MON-4 system classes®MON-4 Collection interval * MON-25 HELP command ® MON-36 /COMMENT qualifier e MON-20 CONVERT command ® MON-33 D 1/0 statistics ®*MON-58 Directing output of MONITOR ® MON-2 /INPUT qualifier* MON-24 Disk file output® MON-11 /INTERVAL qualifier e MON-25 Disk statistics * MON-47 Display data® MON-7 Display output ® MON-6 INITIALIZE command ® MON-37 relationship of qualifier and event® MON-25 specifying for playback ® MON-25 Invoking MONITOR ® MON-2 /DISPLAY qualifier*MON-21 Distributed lock management statistics ® MON-50 L LOCK management statistics ®* MON-60 /ENDING qualifiereMON-22 Index—1 Index MONITOR_SERVER process Recording file log file * MON-42 class header ® MON-98 network startup ® MON-41 class prefix for component classes ®* MON-99 purpose ® MON-41 class records ® MON-97 startup time ®* MON-42 class types ® MON-97 MONITOR ALL_CLASSES command ® MON-38 CLUSTER class record ® MON-124 MONITOR CLUSTER command ® MON-40 data block ®* MON-100 MONITOR DECNET command ® MON-45 DECNET class record®* MON-113 MONITOR DISK command® MON-47 description ®* MON-91 MONITOR DLOCK command ® MON-50 DIGITAL control records ® MON-93 MONITOR FCP command® MON-562 DISK class record®MON-116 MONITOR FILE_SYSTEM_CACHE command ® MON-54 DLOCK class record®* MON-118 MONITOR |0 command ®* MON-58 FILE_SYSTEM_CACHE class record®MON-114 MONITOR LOCK command ® MON-60 file header record * MON-93 MONITOR MODES command ® MON-62 function in MONITOR® MON-10 FCP class record * MON-109 MONITOR PAGE command ® MON-65 1/0 class record®*MON-108 MONITOR POOL command ® MON-67 LOCK class record ®* MON-112 MONITOR PROCESSES command ® MON-69 MODES class record ® MON-104 MONITOR SCS command ® MON-72 node transition record * MON-96 MONITOR STATES command ®* MON-76 PAGE class record * MON-106 MONITOR SYSTEM command ® MON-79 POOL class record®MON-111 Multifile summaries ® MON-12 PROCESSES class record ® MON-100 SCS class record * MON-120 STATES class record ® MON-102 SYSTEM class record ® MON-122 system information record ®* MON-95 version compatibility * MON-11 Network statistics ® MON-45 /RECORD qualifier /NODE qualifier * MON-27 Record type ®* MON-92 MON-29 Nonpaged pool statistics ® MON-67 Restrictions of MONITOR ® MON-3 o S /OUTPUT qualifiereMON-28 Screen format ® MON-7 Output types ®* MON-6 SET DEFAULT command ® MON-82 SHOW DEFAULT command ® MON-83 Summary output® MON-11 P /SUMMARY gqualifier * MON-30 System communication services statistics ® MON-72 Page management statistics * MON-65 System overview statistics ®*MON-79 Process activity statistics ® MON-69 Processor modes statistics ® MON-62 Process states statistics ®*MON-76 \' VAXcluster performance statistics ® MON-40 /VIEWING_TIME qualifier* MON-31 Index-2 VAX/VMS Monitor Utility Reference Manual AA-Z423B-TE e ——— READER’'S Note: This form is for document comments only. DIGITAL will use comments COMMENTS submitted on this form at the company’s discretion. If you require a written reply and are eligible to receive one under Software Performance Report (SPR) service, submit your comments on an SPR form. Did you find this manual understandable, usable, and well organized? Please make suggestions for improvement. Did you find errors in this manual? If so, specify the error and the page number. Please indicate the type of user/reader that you most nearly represent: O Assembly language programmer O O O O Higher-level language programmer Occasional programmer (experienced) User with little programming experience Student programmer O Other (please specify) Name Date Organization Street City State Zip Code. or Country Hereand Tapeg — — — — DoNotTear-Fold — — — — — — — — — — — — — — — — — — | No Postage Necessary if Mailed in the | | I | United States BUSINESS REPLY MAIL FIRST CLASS PERMIT NO.33 MAYNARD MASS. POSTAGE WILL BE PAID BY ADDRESSEE SSG PUBLICATIONS ZK1-3/J35 DIGITAL EQUIPMENT CORPORATION 110 SPIT BROOK ROAD NASHUA, NEW HAMPSHIRE 03062-2698 LI — I B (A P e e A P P A Do Not Tear- Fold Here Cut Along Dotted Line — A VAX/VMS Monitor Utility Reference Manual AA-2423B-TE S —— READER’'S Note: This form is for document comments only. COMMENTS submitted on this form at the company’s discretion. If you require a written reply DIGITAL will use comments and are eligible to receive one under Software Performance Report (SPR) service, submit your comments on an SPR form. Did you find this manual understandable, usable, and well organized? Please make suggestions for N improvement. Did you find errors in this manual? If so, specify the error and the page number. Please indicate the type of user/reader that you most nearly represent: O Assembly language programmer O O O O Higher-level language programmer Occasional programmer (experienced) User with little programming experience Student programmer O Other (please specify) Name Date Organization Street City State Zip Code. or Country | | | | | | | | | | | | | | | Do Not Tear - Fold Here and Tape | — | — No Postage Necessary if Mailed in the United States BUSINESS REPLY MAIL FIRST CLASS PERMIT NO.33 MAYNARD MASS. POSTAGE WILL BE PAID BY ADDRESSEE SSG PUBLICATIONS ZK1-3/J35 DIGITAL EQUIPMENT CORPORATION 110 SPIT BROOK ROAD NASHUA, NEW HAMPSHIRE 03062-2698 — Do Not TearFold Here — — Cut Along Dotted Line —
Home
Privacy and Data
Site structure and layout ©2025 Majenko Technologies
