ULTRIX Guide to Disk Maintenance

This "ULTRIX Guide to Disk Maintenance" (Part Number: AA-ME93C-TE, December 1991, for ULTRIX Version 4.2A or higher) provides comprehensive information for system administrators on using and maintaining disks within an ULTRIX environment.

The document covers four main areas:

1. Disk Organization:

   • Explains disk partitions as logical divisions of a disk for storing different types of files, noting that partitions can overlap (though not recommended for proper setup).
   • Describes the use of disk space for paging and swapping (virtual memory management).
   • Details file systems, primarily the ULTRIX File System (UFS), including their structure (superblocks, inodes) and essential commands for management:
     • newfs for creating file systems (which destroys existing data).
     • mount and umount for accessing and detaching file systems.
     • fsck for checking and correcting file system integrity (requires unmounting the file system).
   • Outlines the default system disk layout, assigning specific partitions for the root directory (/), paging/swapping, and user files (/usr), and discusses system directories like /etc, /bin, /tmp, /dev, and various /usr subdirectories.

2. Managing Disk Space:

   • Instructs on monitoring file system use to ensure adequate free space using:
     • df to check available free space and disk use percentage.
     • du to pinpoint disk space allocation by directory and quot to see per-user block usage.
   • Explains how to implement and verify file system quotas (soft and hard limits on blocks and inodes) using commands like quotacheck, edquota, quotaon, and repquota.
   • Provides methods for obtaining additional disk space, including deleting, archiving, or compressing files, and moving entire file systems or individual files to different partitions or disks. This process often involves backing up data, unmounting, checking integrity, and restoring.

3. Disk Partitioning:

   • Guides users through changing disk partition sizes to accommodate growth or improve performance, primarily using the chpt command.
   • Emphasizes critical pre-partitioning steps: assessing space needs, analyzing performance, reviewing current assignments, and backing up all file systems on the disk to be reconfigured.
   • Details the rules for changing partition tables, such as requiring superuser privileges and not shrinking mounted file systems.
   • Explains how to review current file system assignments (df) and disk partition setup (chpt -q) before making changes.

4. Bad Block Replacement:

   • Describes how the system detects disk errors and performs automatic bad block recovery (controller-initiated, host-initiated, and media replacement strategies).
   • Explains Error Correction Code (ECC) schemes used by disk controllers.
   • Covers bad block detection through terminal messages, the errorlog file, and utilities.
   • Illustrates how to retrieve and interpret DSA and SCSI disk error reports using the uerf command, identifying specific error messages like "Bad Block Reported," "Replace Failure," and "Forced Error Condition."
   • Introduces disk maintenance utilities:
     • radisk for DSA disks (to clear forced error indicators, replace bad blocks, and scan for errors).
     • rzdisk for SCSI disks (to format, reassign bad blocks, and scan for errors).
   • Stresses that manual bad block replacement operations require superuser privileges, single-user mode, and an unmounted disk.

The appendix provides a list of Device Mnemonics used to attach hardware and software devices to the ULTRIX system.