Lost communication

[LinuAIX]

Hi All,

On a p610 that runs BMC's Control-SA Enterprise Security Station using Oracle 8.1.7 yesterday I found I was unable to telnet to the system, or rather telnet got there but the login process failed to start and I would get this:

Trying...
Connected to hostname.domain.com
Escape character is '^T'.

Escape t got me to:

tn>

When I did "open hostname.domain.com" I was informed the connection was already open. I then proceeded to the data center and found I had no login screen on the serial attached console. I switched the IBM3151 off and on, unplugged and plugged back in the serial cable with no success. During this whole time the application was responding and doing its job.

Later that night I ended up pressing the power button. The system then 0c2’d on me (dump). It Then rebooted and went right back to work. Telnet is again alive, the console is just fine. The error report showed nothing out of the ordinary. There was the expected dump:

Failure Causes
UNEXPECTED SYSTEM HALT

Recommended Actions
PERFORM PROBLEM DETERMINATION PROCEDURES

Detail Data
DUMP DEVICE
/dev/dump_lv
DUMP SIZE 51620352

When I did the dd/uncompress I found the dump file was corrupt and thus no help so… I turn to the pros at Tek-Tips and say HELP!

LinuAIX

 

[DSMARWAY]

Hi ,
i would check your system resources i.e. swap space , memory utilisation , whats probably happenend is that your server was heavily utilised , and was slowly running out of resources , eventually your server would have stopped hence power on and off would resolve it.

Check , how much memory you have , and swap space allocated.
i don't use BMC , but BMC could be quite memory intensive and put a load on the system .
Check logs i.e. mail , syslog if you have it set in syslog.conf

HTH