Health Check

[joe_91]

I have to do health check on a server running 11.11. here is what i am planning to do. please let me know what steps i can do further

1. run dmesg
2. check syslog for errors or events
3. check lvdisplay -v on all lvols to see if u have stale
4. swlist -a state |grep -i configured
5. check rc.log

Please let me know

Thanks

Joe

[James R. Ferguson]

Hi Joe:

1. Running 'dmesg' only reports the last critical event. The problem is what you see could be hours or days old. You need to have EMS (Support Tools) in place. Download and install the latest from:

http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B6191AAE

2. Examing the 'syslog' is always a sound action.

3. Checking for stale extents is fine, but having the Support Tools (STM, EMS) in place covers things far more proactively.

4. Running 'check_patches' is a good way to verify the IPD (Installed Product Database) *immediately after* a patch installation.

5. The '/etc/rc.log' is designed only to show the events of the last startup. The '/etc/rc.log.old' shows the last shutdown events. At least the current log should be examined immediately after a reboot or change in the 'init' level. Examination after that point isn't going to provide any useful event insight.

Regards!

...JRF...

[Paul Sperry]

Hi Joe,

I'd also recommend runnig
/opt/sec_mgmt/spc/bin/security_patch_check -r

It will run and create a Security Patch Check Report. I usually run this once a month on all of my systems.

Happy and Healthy computing!!

[joe_91]

I do not have the /opt/sec_mgmt directory. may be only on trusted systems.

Thanks

Joe

[Patrick Wallek]

No, /opt/sec_mgmt has nothing to do with trusted systems.

The /opt/sec_mgmt directory is a part of the security patch check tool, which is available here:

http://h20293.www2.hp.com/portal/swdepot/displayProductInfo.do?productNumber=B6834AA

[Basheer_2]

Hello Joe,

download, install and run shc ( system health check) avail Free from itrc.

[VK2COT]

Hello Joe,

I wrote a Perl script that runs close to 200 tests:

http://www.circlingcycle.com.au/Unix-sources/HP-UX-check-OAT.pl.txt

All that the previous posters mentioned and much more.

I keep updating the script and am very
satisfied because it saved me on many occasions when I was called in to do some work on
servers I never saw in my life before :)

I want, for example to know if a given server:

Has O/S disks on different controllers
Has enough space to extend volumes
Has any hardware problem
Has low-quorum on boot disks
Savecrash setup
LAN setup
SAN setup
Backups
and more

Regards,

VK2COT

[Keith Buck]

I'd also recommend Software Assistant.

https://www.hp.com/go/swa

(redirects to the same link Patrick sent, but with new functionality and better reporting)