Installing and Operating 2.9BSD

This document serves as a comprehensive guide for installing and operating the 2.9BSD release of the Berkeley Software Distribution's UNIX system on a PDP-11. It is adapted from an earlier paper by Bill Joy.

The guide covers:

1. Installation Procedures: Detailed instructions for both full bootstraps (for new installations or upgrades from non-2.8BSD systems) and update procedures (for upgrading from 2.8BSD). It outlines the process of loading the tape bootstrap monitor, creating and initializing a UNIX root file system, and booting the system from disk. It also specifies supported PDP-11 CPUs, disk types (DEC MASSBUS, UNIBUS, AED, DIVA, EMULEX), and tape drives, along with UNIX device naming conventions and the distinction between block and raw devices.

2. System Configuration: Steps for laying out file systems, optimizing disk load, and setting up the /dev directory with special files for peripheral devices. It provides guidance on editing system-dependent configuration files like dtab, fstab, ident, motd, passwd, group, and rc, and configuring terminal lines via /etc/ttys and /etc/ttytype.

3. Kernel Configuration and Compilation: An overview of the kernel's organization, procedures for configuring a new kernel using config, and a detailed list of compile-time options categorized by their impact on system performance or functionality (hardware, global/tunable parameters, general options, file system, performance monitoring, device drivers, system calls, and tuning). It includes specific considerations for automatic reboot and disk monitoring.

4. Recompiling System Software: Instructions for rebuilding system programs, libraries, and the kernel itself, emphasizing the use of MAKE scripts for CPU-specific compilation. It provides guidance on managing local modifications to source code and detailed steps for setting up the mail system (mail, Mail, delivermail) and uucp connections. Brief summaries of additional software packages are also included.

5. System Operation and Maintenance: Procedures for normal system operation, including bootstrapping to single-user or multi-user mode and shutting down the system. It covers analyzing device error messages, performing regular file system checks (fsck), conducting backups (dump, tar) and disaster recovery (restore), moving and resizing file systems, monitoring system performance (iostat, vmstat), adding and managing users, and implementing system accounting (ac, sa).

6. Kernel Magic Numbers: A reference list of "magic numbers" for patching core or executable UNIX binaries, including interrupt vectors and other system variables.

The document stresses the importance of understanding system changes when upgrading and provides contact information for reporting software problems or documentation errors.