Wednesday, April 15, 2009

AIX - System installation and maintenance management

oslevel
Reports the latest installed maintenance level of the system.
oslevel -r : determines the highest recommended maintenance level reached for the current version of AIX.
oslevel -lr 5100-04: lists which fileset updates are missing if after installing ML 04 on 5100-03 the command oslevel -r still shows 5100-03.

alog
Creates and maintains fixed-size log files.
alog -o -t boot : view the boot log (the log that holds boot information).
alog -L : lists the logs defined in the alog database.

errpt
Generates a report of logged errors in the system error log.
errpt -a : displays a complete detailed report.
errpt -c > /dev/console : formats and displays each of the errors at logtime (concurrent error logging) on /dev/console.

errdemon
Starts the error logging daemon errdemon that reads error records from the /dev/error file and creates error log entries in the default system error log /var/adm/ras/errlog.
/usr/lib/errdemon : starts the error logging daemon.
/usr/lib/errdemon -l : displays the path to the system error log file and error log size.
/usr/lib/errdemon -s 2000000 : changes the maximum size of the error log file to 2 MB.

syslogd
The syslogd daemon logs messages from kernel, daemons and system applications using /etc/syslog.conf.
*.debug errlog (add this line to to syslog.conf to redirect all syslog messages to the system error log).
stopsrc -s syslogd : stops the syslogd daemon.
startsrc -s syslogd : starts the syslogd daemon.
refresh -s syslogd : refreshes the syslogd daemon.

errlogger
Logs an operator message.
errlogger new disk added on scsi1 adapter : logs "new disk added on scsi1 adapter" in the system error log.

errclear
Deletes entries from the system error log. Software and operator errors (older than 30 days) and hardware errors (older than 90 days) are removed using crontab.
errinstall
Installs or replaces messages in the error logging message sets of the error log message catalog.

errupdate
Updates the Error Record Template Repository (default file /var/adm/ras/errtmplt).

diag
Menu driven program to run a wide choice of tasks and service aids (diagnostics, hardware error report, format, microcode and bootlist management, ...).
Diagnostics modes:
Concurrent mode: diag is used during normal operation (only devices not in use can be tested).
Single-user mode: run diag after shutdown -m.
Stand-alone mode: boot from Diagnostics CD (press F5 when acoustic beep is heard)
or boot and press F6 when acoustic beep is heard to load diag from hard disk.
if diag returns "diag is not supported on this model" use:
SMS mode: boot and press F1 when acoustic beep is heard, select "test the computer". Some older models use a SMS diskette.

alt_disk_install
Installs an alternate disk with a mksysb install image or clones the currently running system to an alternate disk.
Note: install bos.alt_disk_install fileset to use alt_disk_install.
alt_disk_install -C hdisk2 : Clones the current rootvg to hdisk2.
alt_disk_install -C -b update_all -l /dev/cd0 hdisk4 : Creates clone of the current rootvg on hdisk4, installs a ML on the clone and changes the bootlist to hdisk4.
alt_disk_install -X old_rootvg : Removes the original rootvg from the ODM, after booting from the new alternate disk (you can still reboot from old_rootvg).

nimadm
Performs Alternate Disk Migration (to a new version or release) of AIX using NIM resources.
nimadm -c aix1 -s spot2 -l lpp2 -d "hdisk1 hdisk2" -Y : migrates totarget NIM client aix1, using NIM SPOT resource spot2, the NIM lpp_source lpp2, and hdisk1 and hdisk2 target disks, and agreeing to all required software license agreements for the software being installed (-Y).
nim -o alt_disk_install -a source=rootvg -a disk='hdisk2' -a phase=12 holland : clones a rootvg on client holland to hdisk1, but only run phase1 and phase2 (leaving the /alt_inst file systems mounted).

AIX - Problem Determination

sysdumpdev

Changes the primary or secondary dump device designation in a running system. The default primary dump device is LV /dev/hd6 and the default secondary dump device is /dev/sysdumpnull. A dedicated primary dump device LV /dev/lg_dumplv is created (if sufficient disk space is available) in systems with at least 4 Gigabytes of real memory.

sysdumpdev -l : displays current dump device settings.
sysdumpdev -P -p /dev/hd7 : changes the primary dump device permanently from the default to LV /dev/hd7.
sysdumpdev -e : estimates the dump size (in bytes) for the current running system.
sysdumpdev -L : displays statistical information about the last dump.
chdev -l sys0 -a autostart=true : automatically reboot after a crash (default is false).

dumpcheck
Checks the disk resources used by the system dump and logs in the system error log. Run default by cron at 3:00 pm local time each day.
/usr/lib/ras/dumpcheck -p : requests a dumpcheck. The result is printed to stdout (-p).
/usr/lib/ras/dumpcheck -r : discontinues running dumpcheck (removes the crontab entry).

kdb
Displays system images for examining a dump.
kdb /var/adm/ras/vmcore.0 /unix : starts kdb using the uncompressed dump file /var/adm/ras/vmcore.0 and kernel file /unix.

snap
Gathers system configuration information and compresses the information into a pax file.

snap -a -o /dev/rmt0 : gathers all system configuration information (needs approximately 8 MB space in directory /tmp/ibmsupt) and creates a compressed pax image (snap.pax.Z) of directory /tmp/ibmsupt.

snapcore
Gathers the core file, program, and libraries used by a program to directory /tmp/snapcore (default) and compresses the information into a pax image. The collected information allows debugging and resolving problems within an application.

snapcore -d /tmp/snapcore2 core.xx : gathers all needed information for core dump file core.xx and writes it to directory /tmp/snapcore2/snapcore_32811.pax.Z, where 32811 is the process id ($pid) of the snapcore command.
uncompress -c snapcore_32811.pax.Z | pax : displays the contents of the pax archive.
check_core

Used by snapcore to gather all information about a core dump. The bos.rte.serv_aid fileset must be installed.
/usr/lib/ras/check_core core.xx : displays a list containing the program that caused core dump core.xx and the used libraries.

shconf
Manages the system hang detection parameters for the system hang daemon shdaemon.
shconf -d : displays if priority problem detection and lost I/O detection are enabled or not.
shconf -E -l prio -H : displays the current shdaemon settings.