power supply error!!!!!!!!!!!!!!!!!!!!!!!!!!

[230173]

I'm getting these mails:


Message 1:
From root Mon Jun 3 04:01:41 2002
Date: Mon, 3 Jun 2002 04:01:40 +0200
From: root
To: root
Subject: diagela

A PROBLEM WAS DETECTED ON Mon Jun 3 04:00:35 DFT 2002 801014

The Service Request Number(s)/Probable Cause(s)
(causes are listed in descending order of probability):

805-805: Error Log Analysis indicates the following problem:
Error log information:
Sequence number: 30
power supply FRU: n/a 00-00
Use the SRN to determine the FRU part number.


Does somebody know what this means? The power supply is working so...???

 

[MoshiachNow]

Hi,

1.Which platform ?
2.Once you get such an error in errpt and diag - I would not think twice before replacing the PS.

One of the voltages may be out of range - the user would not know. "Long live king Moshiach !"
h

http://www.noahide.com/movement.htm

 

[Yegolev]

Check for a crontab entry, sometimes diag puts stuff in there. But replace the psu, too. IBM Certified -- AIX 4.3 Obfuscation

 

[Einstein47]

I had a powersupply go out on me almost 2 months ago. There were messages in the errpt as well as the crontab entry that was mentioned.

After the IBM CE came out and installed the new powersupply, the error messages continued. Apparently you need to manually clear the error out of the logs after the new hardware is installed. If you've never done this before, make sure you mention it to your IBM CE when they are on-site.

Also remember that even if you have a redundant powersupply, you should shutdown the system to do the work to avoid any surges or spikes that could cost more than the powersupply. Einstein47
(Love is like PI - natural, irrational, endless, and very important.)

 

[Yegolev]

Yep, the CE won't "fix" your crontab unless you tell him to, and then he'd need your root passwd unless you log him in.

Some machines (such as the H50) have a redundant psu but if you pull one out for maintenance you only have five minutes to put the new one in. Your CE should know that, but there was a guy who took five minutes and three seconds once.... IBM Certified -- AIX 4.3 Obfuscation

 

[TysonG]

We recently just got a new 7026-6H0 and it was logging similar errors right from the get-go. All 4 power supplies were working fine, so I thought the error was bogus. I didn't actually check the crontab, but I did clear the error. I did a "diag", then task selection, then log a repair action. I picked sysplanar0 as the device and pressed enter. Voila, no more diagela messsage.

I'm pretty sure this is just an easier way to clear the crontab of these specific messages, but it does make the messages go away for sure. If the messages do come back, one of your power supplies truly could be dead. Power supplies sometimes stay lit up even when they aren't working properly.

 

[Yegolev]

Not sure that the Repair Action does anything to the crontab, but it certainly is the best way to get diagela to be quiet without clearing the error log. IBM Certified -- AIX 4.3 Obfuscation

 

[ggitlin]

Try upgrading the system firmware and service proc microcodes. One of the patches addresses power supply problems.
Here is the site for F50/H50:

http://techsupport.services.ibm.com/server/mdownload/7026H50F.html

 

[230173]

Well, I've cleared the errpt and it stopped giving me these error mails!!!!

Thnx