Tonneson DokuWiki

• Virtual processor: 0.1-1.0
• SMT on Power 7: 1 vCPU = 4 logical
• DLPAR can only consume physical CPU equal to entitled virtual CPU
  • example: 2 vCPUs can consume up to 2 physical cores max

AIX: (Advanced Interactive eXecution)

smitty -x:creates scripts, no action

don't run cfgmgr on a large system w/o switches

z series = mainframe x series = x86 i series = as/400 (deprecated now just 'power') p series = power (deprecated now just 'power')

Find size of DIMMs (memory) from AIX command line

lscfg -vp | egrep 'Memory DIMM|Size'

Add paging space

lsps -a
lsvg rootvg | grep -e "FREE PP" -e "PP SIZE"
chps -s 8 hd6
chps -s 8 paging00
lsps -a

Display effective system attributes

lsattr -E -l sys0

which_fileset In a situation where the AIX server you are working on does not contain the AIX command(s) you need, just use the which_fileset command to identify which fileset needs to be installed.

Example:

$ which_fileset dbx
/usr/bin/dbx -> /usr/ccs/bin/dbx        bos.adt.debug 6.1.0.0
/usr/ccs/bin/dbx                        bos.adt.debug 6.1.0.0
$ which_fileset lscfg
/usr/sbin/lscfg                         bos.rte.diag 6.1.0.0

Then you know which fileset needs to be installed.

You can Install the missing fileset(s) and you're back in business. In the example above, it would be bos.adt.debug and bos.rte.diag (if they were missing).

Note: The which_fileset command needs the bos.content_list fileset installed to work. This should be all most AIX systems.

Create backup of /etc/inittab file:

cp /etc/inittab /etc/inittab.YYYYMMDD

Modify /etc/inittab (use perfect syntax) Refresh init process:

telinit q

There is a known problem with mounting ISO images on the AIX platform. To workaround the problem, complete the following instructions:

• Obtain the size of the image.

ls -al

• Identify the nearest multiple of 128 MB that will provide enough space for the image.

• Make a logical volume of this size. Ensure that there is enough space on the physical volume. For example, hdisk0. If the command fails, increase the volume size by a multiple of 128.

mklv -y cdlv -s n -L /dev/cdlv rootvg 384M hdisk0

• Use the dd command to create a pseudo-device. Ensure that the partition has enough space for the pseudo-device. In the following example command, the pseudo-device is /dev/cdlv. Note that this command may take a long time and will create two dd processes.

dd if=/opt/software/iso/fim_console_rios_aix_5.iso of=/dev/cdlv

• Mount the device like a CD-ROM in AIX. Ensure that the mount point exists.

mount -v cdrfs -o ro /dev/cdlv /mnt/iso

• Change directory into /mnt/iso.

If there is no filesystem created in /etc/filesystems for a cd device, create one:

/usr/sbin/crfs -v cdrfs -p ro -d'cd0' -m'/cdrom' -A'no'

Then mount cdrom

mount /cdrom

List the defined log types:

alog -L

View the boot log:

alog -o -t boot

View the console log:

alog -o -t console

This is just an example to show the steps involved in re-configuring a VSCSI client adapter on a DLPAR in an environment with dual VIO servers and MPIO at client LPAR. Be sure you know exactly what you are doing. This can be a risky operation if you don't understand everything.

• List missing paths with “lspath | sort” command.
• If there is one path to each disk, then this can be a dynamic change.
• Remove missing paths on LPAR with command “rmpath -dl hdiskX -p vscsiX”
• Remove bad vscsi devices with command “rmdev -dl vscsiX”
• Dynamically remove client virtual scsi adapters from HMC
• Run “cfgmgr;lsdev| grep vscsi” to ensure the vscsi adapters are gone.
• Fix and recreate virtual scsi adapters from HMC
• Run “cfgmgr; lsdev | grep vscsi; lspath | sort; lscfg -l vscsiX” to ensure proper configuration
• On HMC, rename active, default profile to something like “foo_DO_NOT_USE_OLD_CFG_bk1”
• On HMC, save current configuration to what the old default, active, profile was (i.e. foo_default)
• On HMC, set default configuration to newly saved configuration

Error: 0514-604 Cannot access the {CuDv} object class in the device configuration database.

odmdelete -o Config_Rules -q “rule = /etc/methods/darcfgrule”

Sometimes this occurs after uninstalling the Shark filesets or doing mksysb restores which leave a hanging piece of crud in the ODM. Be careful here!