changing 32bit kernel to 64bit kernel

how would you change a 32bit kernel to 64bit. In aix5L the procedures are:

1. smitty load64bit

This will take you to the following menu :

Enable Now
Enable/Disable at System Restart

2. Select enable now and then select enable 64 bit env. at system restart.
3. A line is added to inittab to load this automatically at system reboot.

If you are not sure, try this command:

To switch from 32-bit mode to 64-bit mode run the following commands, in the given order:

1. ln -sf /usr/lib/boot/unix_64 /unix
2. ln -sf /usr/lib/boot/unix_64 /usr/lib/boot/unix
3. bosboot -ad /dev/ipldevice
4. shutdown -Fr
5. bootinfo -K (should now show 64)

To switch from 64-bit mode to 32-bit mode run the following commands, in the given order:

1. ln -sf /usr/lib/boot/unix_mp /unix
2. ln -sf /usr/lib/boot/unix_mp /usr/lib/boot/unix
3. bosboot -ad /dev/ipldevice
4. shutdown -Fr
5. bootinfo -K (should now show 32)

Now, the biggest question would be, whu 64bit?

The advantages of a 64 bit kernel in memory addressing, maximum file size, maximum partition size, in the LVM if you use it and of course it faster.

FYI .In AIX4.3 the kernel is only 32-bit.

pwc2id5hr3

IBM Power servers most reliable in new survey

Windows Server machines most improved, but still suffer high downtime By Jon Brodkin , Network World , 07/14/2009

IBM's Power servers topped a list of most reliable x86 and Unix machines in a new survey, clocking in at only 15 minutes of unplanned downtime peryear.

Linux distributions running on x86 servers also performed well, as did Sun's Sparc machines and

HP's Unix boxes. Windows Server machines performed worse than most competitors, with two

to three hours of downtime per year, but have still improved dramatically over previous surveys.

"Ten to 15 years ago, there was a lot more downtime [for all types of servers]," says Laura

DiDio, lead analyst with Information Technology Intelligence. "Both the hardware and operating

system software has gotten much better."

ITIC picked 15 of the most popular server hardware and operating system combinations, and polled Clevel executives and IT managers at 400 organizations across 20 countries about unplanned downtime, patching and other indicators of reliability.

IBM's Power servers with the AIX Unix operating system suffered an average of 15 minutes of

unplanned downtime per year in the most recent survey, down from 30 minutes in the 2008 survey. IBM also came in first place for average time to patch a server (11 minutes per patch) and for lowest number of outages per year.

"For the second year in a row, IBM AIX UNIX running on the Power or "P" series servers, scored the highest reliability ratings among 15 different server operating system platforms -- including Linux, Mac OS X, UNIX and Windows," DiDio writes in a blog post. The second most reliable servers in terms of downtime were customized versions of Novell SuSE Linux

running on standard x86 hardware, clocking in at 17.4 minutes of downtime per year. Un-customized Novell SuSE Linux machines had 54 minutes of downtime.

Linux distributions such as TurboLinux and Mandriva on standard x86 hardware suffered 31.8 minutes of downtime a year, while Sun Solaris on Sparc servers suffered 35.4 minutes downtime a year. HP 9000 servers running HP's Unix operating system came in fifth place with 36 minutes of downtime per year, while HP also took seventh place with Integrity servers (39 minutes downtime).

Apple's G4 Mac servers with the Mac OS X operating system came in sixth with 37.8 minutes

downtime. "The survey respondents indicated that Apple products are extremely competitive in an enterprise setting," DiDio writes.

The worst performers were open source Linux distributions such as Debian, with more than four hours unplanned downtime per year. The next-worst were Windows Server 2003 on Intel-based hardware (three hours of downtime) and Windows Server 2008 (nearly two and a half hours downtime). But Windows Server systems also posted the biggest improvement, with a 35% reduction in downtime since2008.

Ubuntu-based servers seemed to go backwards, from one hour of downtime in 2008 to one hour, 41 minutes in 2009. That is more a function of the type of user attracted to Ubuntu, however, DiDio says. "Ubuntu is very leading edge. Ubuntu users do a lot of experimentation" and can thus expect some downtime, she says. DiDio attempted to measure reliability of most popular enterprise servers, she decided not to include mainframes, which probably would have taken the top spot.

"Mainframes are in a class by themselves," DiDio notes. "Whatever you say, you're not taking the mainframe down. It's like the Rock of Gibraltar."

This story appeared on Network World at

http://www.networkworld.com/news/2009/071409-ibm-power-servers.html

IBM REDBOOKS JULY RELEASED

PowerHA for AIX Cookbook

Revised: July 6, 2009

http://www.redbooks.ibm.com/redpieces/abstracts/sg247739.html?Open

p6-520 & p6-550 System Builder

Revised: July 7, 2009

http://www.redbooks.ibm.com/redpieces/abstracts/sg247765.html?Open

PowerVM: Intro & Configuration

Revised: July 10, 2009 ISBN: 0738485306 394 pages

http://www.redbooks.ibm.com/abstracts/sg247940.html?Open

High Availability Solution for IBM FileNet P8

Revised: July 7, 2009

http://www.redbooks.ibm.com/redpieces/abstracts/sg247700.html?Open

ServeRAID Adapter Quick Reference

Revised: July 7, 2009

http://www.redbooks.ibm.com/abstracts/tips0054.html?Open

Brocade 10Gb CNA for IBM System x

Revised: July 6, 2009

http://www.redbooks.ibm.com/abstracts/tips0718.html?Open

QUOTE: he who knows all knows nothing

TSM commands (quickie)

Some commands gathered during my stint as a tsm administrator. a very short stint. Anyways assumptions are:
library name is 3494lib.
physical tape media is B00001
for finer details just add f=d at the end of the commands. e.g. q mount f=d..
not all commands works and be careful when executing especially when updating status of private to scratch.

1. to update tape in private status to scratch status
tsm: ADSM> update libvol 3494lib B00001 status=scr

2. to check status of tape last written/read
tsm: ADSM> q vol B00001 f=d

3. to check the content of tape
tsm: ADSM> q con B00001

4. check which tape is mounted for during operation
tsm: ADSM> q mount

5. to check whether the session of tsm has started
tsm: ADSM> q session

6. to check the sotrage pool operation
tsm: ADSM> q vol * stgpool=3494lib

7. to erase the content of data on tape and turn it to scratch
tsm: ADSM> del vol B00001 discarddata=yes

8. To update the path to drive1:
TSM>upd path ADSM 3590DRIVE1 srct=server destt=drive online=yes libr=3494lib autod=yes

9.To check the volume mounted:
TSM>q mount

10. to update the path
tsm: ADSM>update drive 3494lib 3590drive3 online=yes

11. to check the activity log from 3 days ago until today
tsm: ADSM>q actlog begind=today-3

12 to check status of path in finer details
tsm: ADSM>q path f=d

13. to check schedule of works in TSM
tsm: ADSM> q sched t=a

HACMP commands

Sometime i forgot certain commands when trying to figure out HACMP stuffs. So i look through Mr Google and find some that i think might be useful. Will add some more.

clstat - shows cluster state

cldump - SNMP-based tool to show cluster state

cldisp - perl script to show cluster state

cltopinfo - list local view of cluster topology

clvt -

clshowsrv - local view of cluster subsystem

clfindres, clRGinfo - locate resource groups for displaying status

clcycle -

cl_ping - cluster style pinging

clgetactivenodes - show nodes that are active

get_local_nodename - get local node name

clconfig

cldare - syncro cluster, be careful using this command

cllsgrp - list the resource group

cllscf - list network config of a cluster

clshowres -

cllsif - lists network interface information

cllsres - lists resource group info

lssrc -ls topsvcs - hearbeat info

lssrc -ls clstrmgrES - lists cluster manager internal state

clver -Nv -

clharvest_vg -w -

notice that some commands have no explaination. I haven't tested the command and therefore wouldn't know the effect. Try it yourself. Never know,eh? Just be careful. Somebody told me to never forgets to do system backup or 'mksysb' before testing and doing stuffs on AIX.