HPE 9000 and HPE e3000 Servers
cancel
Showing results for 
Search instead for 
Did you mean: 

Single bit error (SBE) event.

 
SOLVED
Go to solution
Alberto Hurtado
Frequent Advisor

Single bit error (SBE) event.

Good morning !
I was received tree msgs. with diferent severity (4000, 4100, 4300) of single bit error. What can i wait in a shot time ?? I must replace this dimm of memory for this reason ?

Thank's in advance
6 REPLIES 6
Sanjay_6
Honored Contributor
Solution

Re: Single bit error (SBE) event.

Hi,

Check your memory using this,

echo 'selclass qualifier memory;info;wait;infolog' |cstm >/tmp/meminfo.txt

Then vi this file and you should see all the modules listed over there with their sizes. You'll also see if there were any errors and the modules for which there were errors. One or two errors far spaced would not be a problem, but otherwise it is a problem even if it is a single bit error.

We always try and get the module replaced rather than waiting for a error to happen again. Can't wait for a production server to go down on its own.

Hope this helps.

Regds
Jeff Schussele
Honored Contributor

Re: Single bit error (SBE) event.

Hi Alberto,

Single bit errors are common & if not in locked memory the system will deallocate the page as a normal course of events. If you can reboot the system it will deallocate the page if indeed it has a bad spot. Sometimes HP will not replace the DIMM if it only has a few bad bits. They'll insist you reboot to deallocate the page - but sometimes they will replace the DIMM. It just depends. You should check the PDT (Page Deallocation Table) using the stm commands. IF there are quite a few deallocations on the same DIMM they will definitely replace it. If there's only one or just a few, they usually won't.

Rgds,
Jeff
PERSEVERANCE -- Remember, whatever does not kill you only makes you stronger!
Alberto Hurtado
Frequent Advisor

Re: Single bit error (SBE) event.

I was looking the board of memory
Board Error Address Error Type Page Count
------------ ------------------ ---------- --------- -----
MB6R/1S1 0x00002d01994f01a8 Single-Bit N/A 75

And you are in the correct, but in reference to Jeff, the PDT there are not deallocations on the same DIMM. i Think first reboot my server HP V2250 in PRD.

Jeff Schussele
Honored Contributor

Re: Single bit error (SBE) event.

Hi (again) Alberto,

Now that I know you're dealing with a V-class here, I should mention that there is a known issue with STM (Support Tool Manager) and V-class. Here are the patches that HP told us to apply depending on the release level of STM you have.

The following patches have been released to correct the single-bit memory error problems you have been experiencing on the V-classes.

PHSS_29343 - s700_800 11.11 Support Tool Manager Mar 2003 Patch

PHSS_29344 - s700_800 11.11 Support Tool Manager Jun 2003 Patch



They are in official release and are available via ITRC.


The problem we observed was that the address was being miscalculated & being reported as in locked memory when in actuality it wasn't. This caused the address to *not* be deallocated when in fact it could have & should have. The only way it would be deallocated was on reboot. And in the meantime we'd get dozens & dozens of alert e-mails. The patch did solve the problem.

Rgds,
Jeff
PERSEVERANCE -- Remember, whatever does not kill you only makes you stronger!
Alberto Hurtado
Frequent Advisor

Re: Single bit error (SBE) event.

Thank's Jeff, tonight i will aplly a reboot.
And we will see that he is what happens, while i would are downloading the patches. greetings

Alberto
Andrew Merritt_2
Honored Contributor

Re: Single bit error (SBE) event.

Yes, there is a known problem on V-class systems which may be the cause of what you're seeing.

Before you can apply those particular patches, you'll need to upgrade the OnlineDiags installed. The attachment you posted shows you have A.25.00, which is the March 2001 release of OnlineDiags, now a very old release.

If you upgrade to the September 2003 (A.42.00) or later OnlineDiags, you won't need to install a patch to get that particular fix, since the fix was included with that release (but you should still install the latest patch for that revision of OnlineDiags for other reasons).

Latest releases can be seen at http://www.docs.hp.com/hpux/onlinedocs/diag/stm/stm_upd.htm#table

http://www.docs.hp.com/hpux/onlinedocs/diag/st/st_web.htm tells you where you can download from.

Andrew