Log files or trouble shooting tips

[bcny33]

Hello,

When my server(sparcserver20) crashed or actually just kind of locked up.. It went back to normal after a reboot... But where should I look to trouble shoot why it happened??... When I looked under the /var/adm/messages file it just basiclly gives me the info in regard to booting up. I Hope this is not a stupid question but I'm not understanding how to Pin Point the problem when something like this happens in Solaris... For example there have been a few mornings in the past when I come into the office and I notice our Authenication Solaris server is on the White screen of death (Ok prompt)... I would understand if I lost a hard drive or something like that but my Hard drives seem to be fine.... Any input would be much appreciated... I'm still learning so bare with me...

Thanks
bcny

 

[spamly]

Try the command: dmesg
This will report many hardware errors.

Also, can you tell if the systems panicked? What is displayed on the console before the ok prompt? There's a chance that the server reset and your prom variable "auto-reboot?" is set to false. If you changed this to true (See man eeprom) the system may automatically boot back up after something like this happens.

 

[djr111]

The best way to capture the best logs/errors is to have a console window open there are messages that will go to console that sometimes ncever make it to /var/adm/messages. If you have a network console it is the easiest, but if not, keep a PC hooked up to the console using hyperterminal. or any of the other ways. You can run a console window within the server, but if one of your issues is that it will crash, this will not do to much since the console window will die once the OS panics.

hope this helps