DL380 Overheating Issues

[fox1977]

Hi folks,

Just wondering if you can help me out. We have ahad two of our live database servers go down over the weekend as a result of the pair of them shutting themselves down as they were over heating. Here are the entries in the event logs:

Server 1
---------
Event Type: Warning
Event Source: cpqasm2
Event Category: None
Event ID: 4120
Date: 30/06/2007
Time: 18:57:15
User: N/A
Computer:
Description:
The temperature (thermal sensor #5) of the system has exceeded the
threshold.

From the HP event log:
System overheating Zone 5 - Location: Power Supply Unit Temp 51c

Server 2
----------
Event Type: Warning
Event Source: hpqilo2
Event Category: None
Event ID: 7
Date: 30/06/2007
Time: 18:50:15
User: N/A
Computer:
Description:
Thermal sensor 1 has exceeded its threshold temperature.

From the HP event log:
System overheating Zone 6 - Location: IO Board Temp 66c
--------------

Anyone else ever experienced any similar problems?

I know there is a heat issues but are there any fans in these zones that I can check on to see if they are still working?

The servers are located in a datacentre at the other end of the country and I am not familiar with what the setup is inside a DL380.

ANY advice gratefully received as I need to get to the bottom of the causes of these shut downs. We have 12 servers in the cabinet and these were the only two to report any problems.

Thanks in advance.

[KarloChacon]

Dl380 generation ?

regards

[fox1977]

Sorry should have posted that!

ProLiant DL385 G2

[KarloChacon]

it could be fan issue check your fans:

and look for system overheating

http://h20000.www2.hp.com/bc/docs/support/SupportManual/c00300504/c00300504.pdf


Fan problems
General fan problems are occurring
Action:
1. Be sure the fans are properly seated and working.
a. Follow the procedures and warnings in the server documentation for removing the access panels
and accessing and replacing fans.
b. Unseat, and then reseat, each fan according to the proper procedures.
c. Replace the access panels, and then attempt to restart the server.
2. Be sure the fan configuration meets the functional requirements of the server. Refer to the server
documentation.
3. Be sure no ventilation problems exist. If you have been operating the server for an extended period
of time with the access panel removed, airflow may have been impeded, causing thermal damage to
components. Refer to the server documentation for further requirements.
4. Be sure no POST error messages ("POST error messages and beep codes" on page 92) are
displayed while booting the server that indicate temperature violation or fan failure information.
Refer to the server documentation for the temperature requirements for the server.
5. Access the IML to see if any event list error messages (on page 124) relating to fans are listed.
6. Replace any required non-functioning fans and restart the server. Refer to the server documentation
for specifications on fan requirements.
7. Be sure all fan slots have fans or blanks installed. Refer to the server documentation for requirements.
8. Verify the fan airflow path is not blocked by cables or other material.

regards

[Pieter 't Hart]

have you checked the insight management home page for the current temperature?
it's important to also check the other servers in the rack and look if the temperature is climbing from bottom to top in the rack.

let someone at the remote site make a picture of the situation.
are alle the disk positions fitted with disks or blanks?

there may be a air flow issue outside the cabinit wich causes the intake air temperature of the server to be too high!
there may be another component in this cabinet(LAN or SAN switch) wich blows out air from back to front?
are there any cabinets in front of this one ?

[fox1977]

Hi there,

Thanks for all the prompt advice folks. I've been chasing support all day to see if they expereinced any problems with their air con. They reported that all of the fans in the two servers are working ok.

I am now looking at getting something to monitor the temperature in the servers.

Where abouts in the HP Insight Management home page does it report the temp? Is this the software that comes with the server. I have the System Management web application and the Insight Diagnostics but I cannot see anything in either that relates to temperature.

Thanks again for all the help

[Thomas Olsen_1]

Hi

From System Management Homepage (part of Proliant Support Pack) under System, click on Environment to display current temperature and status of fans.

/Thomas

[radnam]

HEy i have a interesting topic that i read on one of the forums may be of some interest to you guys:

http://www.eukhost.com/forums/f43/reasons-servers-go-down-5658/

[Michael Nordsven]

Temperature sensor #1 is more than likly the SID board on the front of the server.   That is if it is Generation 9. It has one screw and is easily replaced.