High temperature issue

[affin]

Hi Guru's,

Our environment is composed by many HPUX machine (rp5470,l2000 etc...). Tonight due to cooling problem the temperature has growth very much and all the machine have auto-powered off. All the machines except an RP 4440. On syslog no message indicating the temerature issue, but now we have the temperature led turned on and orange. Could someone explain me what happened and how I can turn off that led (we also used GSP command ShowLogs without success)??


Many thanx
AP

[Sridhar Bhaskarla]

Hi,

'envd' daemon monitors the environmental conditions and takes action according to /etc/envd.conf. However, it does log messages into syslog unless it is configured not to in envd.conf. Check your envd.conf file. It should have 'reboot -hq' associated with emergency conditions.

-Sri

[affin]

hi,

these are the lines into the envd.conf:

OVERTEMP_CRIT:y

OVERTEMP_EMERG:y
/usr/sbin/reboot -qh

FANFAIL_CRIT:y

FANFAIL_EMERG:y
/usr/sbin/reboot -qh


now our issues are two:

1- why this machine has not rebooted while the all others yes(no message appear in syslog)?

2- how to turn off the orange temperature led.


Regards,
AP

[Mark Greene_1]

1. is envd running?

2. the sensor that activates the led may be fried and in need of replacement.

mark

[affin]

1. Yes is up and running.


But how is it possible that other machines doesn't rebbot?
Also is there an hp tool for monitoring temperature and sensor status?

Thanx
AP

[Scot Bean]

The other machines did reboot. The reboot "-h' option powers down the machine to protect it when too hot.

On the rp4440, from GSP try the PS command to see what the current temperature status is.

[Sridhar Bhaskarla]

Hi,

Looks like the sensor didn't sense it properly on your RP4440. You can check it from it's GSP's 'ps' command. If you want, you and look at previous logs to see if there was any condition. Type "sl" and selecte "Error" logs. Setup the filters and I believe it is the "Protection Domain" where you will see environmental conditions.

-Sri

[Scot Bean]

Different models can have different threshholds for temp. The rp4440 could be more heat tolerant, or, it could have a problem.

Also, I assume OnlineDiag is installed. EMS should send email to alert on these conditions. Default is to send email to 'root' user on the machine, but you can configure this to send email to you.

[Dave Hutton]

Like Scot mentioned. Online Diags/EMS will notice a problem. We had an issue a few months ago.

Both syslog.log and roots email did have a notification. For more detailed info I believe I had to go into the GSP logs.

We ended up buying a separate appliance with some temp probes to send events to ITO to let us know if the temp gets too high. For us our reaction time is less then 20 minutes, since the temp grows extremely fast once the chillers aren't working.

[affin]

The PS coomand is indicating that's all well. No lines on syslog no line into the mailbox of root...