We have a 2600 with 2Gb RAM (4x512Mb modules). We added an additional 2x 1Gb RAM modules and all seemed ok. A few hours later the server crashed. Openmanage reported a problem with the new memory & diagnostics wouldn't run as single bit ECC was disabled. Took the memory out & it was ok again. Tried the memory in a second 2600 and it tested all ok on the extended RAM diagnostics. Put it back in the first server (in a different bank) and it failed again. The server passes all the diagnostics, but fails when the 2x 1Gb modules are fitted - any suggestions? (Both servers are BIOS A14.)
All the RAM is the same speed rating and ECC registered. Yep - I tried the 2x 1Gb sticks by themselves with the same error message. (In the second server the sticks are fine by themselves...)
Since you have Open Manage Server Administrator loaded, you should be able to re-enable the Error Correction Count (ECC) by following the below instructions. The ECC is sometimes disabled during Single Bit Errors (SBE) or Multi Bit Errors (MBE).
The command to clear the error count is dcicfg32 command=clearmemfailures Depending on the version of openmanage that is installed you may find dcicfg32 in a variety of locations. I would suggest opening a cmd prompt to the root of c: and running dir /s dcicfg32.exe to locate the utility. In some versions of Open Manage, this is located at c:\program files\dell\openmanage\omsa\bin\. In others, it is located at c:\program files\dell\sysmgt\omsa\bin\.
Nick St
1 Rookie
•
106 Posts
284
0
Posted February 5th, 2008 17:00