rx2620 shuts down after 20-30 minutes.

[Tim Nelson]

The only thing I see in the book is there is a statement about a safety interlock between the front panel and the power supply.

http://docs.hp.com/en/2411/rp24xx_customer.pdf
Page 161

[Torsten.]

Tim, it's an Integrity rx2620, your doc is about the HP9000 A-class, aka rp24xx.
Totally different system.

The interlock on the A-class will immediately switch off the power supply - but it's a different server ...

;-)

I have sometimes problems with names too ...

[rsussman]

Hi,
I have tried everything suggested in this thread, (thanks for the suggestions), but nothing seems to solve the problem.
Does anybody have any other ideas?

Rick

[Grayh]

Hi Rick,

This is a known issue on the rx2620/rx4640 servers.

If the BMC firmware on the server is less then 3.48, the server goes down.A system board with the minimum BMC firmware version of 3.48 (AB331-69201 or later) is the only system board that you can use in an Office Friendly rx2620 server.

The latest version of BMC is 4.27.It is recommended to upgrade to 4.27 if not already done from the following link:-

http://h20000.www2.hp.com/bizsupport/TechSupport/SoftwareIndex.jsp?lang=en&cc=us&prodNameId=441910&prodTypeId=15351&prodSeriesId=447331&swLang=13&taskId=135&swEnvOID=7

If the above dosent resolve the issue send the following logs to analyse:-

/ver/tombstones/ts99
#tail -1000 /var/opt/resmon/log/event.log
#cat /etc/shutdownlog
#cat /etc/OLDshutdownlog
#tail -n 1000 /var/adm/syslog.log

And also the following from MP to analyse the issue:-
MP>SL>E>D

MP>SL>F>D

MP>CM>PS

MP>CM>DF

MP>CM>SYSREV

Regards,
Mohan

[rsussman]

Hi,
SYSREV gives me:
MP FW: E.03.30
BMC FW: 04.04
EFI FW: 05.48
SYSTEM FW: 04.27

PS
Eeverything is normal

I am considering re-installing the OS. I have 2 other identical 2630s in the same rack not experiencing this problem.

Thanks...Rick

[Khairy]

hi,

i don't think reinstalling the OS will solve the issue as its its clearly indicate a temperature problem. Check system event log.

All integrity servers are equip with thermal logic management which constantly monitor the temperature and will initiate appropriate corrective action such as initiating system shutdown when temperature reaches critical levels. This is to avoid the server from overheating as we're all know that overheating may cause futher damages to electronic components in the system and I/O boards.

Check for current temperature where the server is located. Make sure proper ventilation is achieve if the server is rackmounted.

Rgds

[Sameer_Nirmal]

The period of 20-30 minutes does match with the pre-defined threshold in the BMC firmware. So I would think this has to do with the chassis open sensor for sure.

This I would do to verify ( ensuring chassis cover is placed properly). Once the system is powered on , monitor MP FPL log , look for the first chassis open event and events thereafter. If there is chassis open error immediately after power on , the event log shows multiple entries until it reach the threshold and shut down the system, it would mean that the sensor is faulty and would require system board replacement. I would also take a look at the MP-CM-DF result to see if the sensor entry is listed even though it may not be a FRU.

This sensor is optical sensor. I am not sure about it's exact location it maybe around
the area between memory slots and PCI slots on the system board.

[Murat SULUHAN]

Hi Rick

0x2047A0E8A50202B0 020EA87002120300 --> definitely sensor related problem not fan related.

You may look at OS event.log and other System Event Logs in MP. You can also open a ticket local HP support

Best Regards
Murat

[rsussman]

Hi,
The 20-30 minutes was a guess. The system stays up longer than 30 minutes.

Thanks!

[Grayh]

Please post the following logs to analyse the issue:-

o/p of the attached script

Procedure for script execution.

Use the binary mode to transfer the files. In case you use ftp mode , Copy the script in the working directory and execute the following command.

# sh getsysinfo.sh -t

The output of this script would be 'sysinfo.tgz' in '/tmp' directory.

Could you also send us the following MP logs for each of the servers for a helthcheck :-

MP>SL>E>D

MP>SL>F>D

MP>CM>DF