Deallocation for a full PDT table Replace the DIMM(s) that were deallocated

[Muthyala]

Hi

I am getting above error on my RP7420 server running hp-ux 11iV1.This error started after i updated the server with Quality Pack Depot for 11i v1, December 2007.

IPMI event hex: 0x6480042e00e00000 0x00ffff00ff2aff74 Time Stamp: Wed Jul 18 01:31:09 1928 Event keyword: MEM_DIMM_ERR_DEALLOCATED Alert level name: Warning Reporting vers: 1 Data field type: Physical location
Decoded data field: FRU Physical Location: 0x00ffff00ff2aff74
FRU Source = 7(memory)
Source Detail = 4(SIMM or DIMM)
Cabinet Location = 0
Cell Location = 0
DIMMS Rank/Echelon = 2a


My CSTM O/P

Cab/Cell in Slot 0/0
==========================================================
DIMM Slot: 2a
Error Type: Single/unconfirmed: single-bit error that
could not be confirmed as either soft or hard.
Page Status: Active: page is active and can be used.
Bit Num: -1
Logged By: Memlogd
First Detected: Mon Mar 3 11:31:39 2008

Last Detected: Mon Mar 3 11:31:39 2008

Error Count: 1
Error Addr: 0x82797fa00
Rank: 8
Memory Sub System:2
Syndrome: 0x3a9
==========================================================

Cab/Cell in Slot 0/0
==========================================================
DIMM Slot: 2a/b
Error Type: Single/unconfirmed: single-bit error that
could not be confirmed as either soft or hard.
Page Status: Pending: page could not be obtained.
Bit Num: -1
Logged By: PDT
First Detected: Mon Mar 3 11:31:37 2008

Last Detected: Mon Mar 3 11:31:37 2008

Error Count: 0
Error Addr: 0x8281ffa01
Rank: 0
Memory Sub System:0
Syndrome: 0
==========================================================


Let me know if i need to replace the memory module.

TIA
Sri

[Tim Nelson]

One of two things: Either diags got better at detecting or it is now worse at detecting things.

Why not place a call to HP just to be sure ?

I would typically monitor through xstm expert tool. If the errors keep growing then place a call.

[Patrick Ware_1]

In our company, 1 error does not warrant a call. If we keep getting the errors, then a call is placed before the PDT gets full, and it is fixed. Monitor to make sure no more errors show up today.

[Andrew Rutter]

hi,

did you update the diag tools aswell or just the hardware bundles?
Also is the EMS and STM tools upto date? or older versions installed?

But it does seem as though you have single bit errors appearing, but these are correctable at this time. If these turn into double bit errors then the server wil crash,with a HPMC.

Monitor the memory and do a few tests on the memory with STM and see how it goes, but if its on support then you may aswell get it swapped out

Andy

[Muthyala]

Hi Andy

Can i go ahead replace the said DIMM's with new ones.

TIA
Sri

[Andrew Rutter · Accepted solution]

hi,

yes of course you can as long as you can get the downtime! try and schedule it in.

Better to do it while you can and have choice rather than wait for server to crash and have headache

you should be able to see how the dimms are populated and the sizes through either STM or rebooting the server and at the BCH menu

Andy

[Torsten.]

If the diagnostic sees any problem, it is trying to free up the memory area and test it.

But here it can get the area:

"Page Status: Pending: page could not be obtained"


I would consider to update the diagnostic version, reboot the server and monitor the status.

[Ron Irving]

Sorry so late in aswering.

cstm would be able to accuratly detail the memory errors, and which DIMM(s) to replace.

In my experience, if memory is installed in pairs, it should be replaced in pairs. Same with quads.

Don't forget to reset PDT in the BCH Service menu when you startup again, (Service>PDT Clear)

Hope this helps.

ron