[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: 2.6.9-55.ELsmp EDAC Errors



I got these on Sun x4100's with Opteron 280 and 285 procs. I did not get them with 4.4 (42 version of the kernel), but did with 4.5 (55 version of the kernel). I found that Sun had a firmware patch that was required to work with the new 4.5 version of Red Hat. You might check with your hardware vendor. After applying the patch, I have received no more of the errors, and they were happening constantly before the firmware upgrade.

Hope this helps.

-Troy

Mike Hanby wrote:
Howdy,

I just installed RedHat 4.5 on 64 identical nodes (Dual AMD Opteron 242
with 2GB ECC RAM).

4 of the nodes are logging errors similar to:

Jul 25 11:07:30 node1.local kernel: EDAC k8 MC0: extended error code:
ECC chipkill x4 error

Jul 25 11:07:30 node1.local sshd[3027]: error: Bind to port 22 on
0.0.0.0 failed: Address already in use.

Jul 25 11:07:31 node1.local kernel: EDAC k8 MC0: general bus error:
participating processor(local node response), time-out(no timeout)
memory transaction type(generic read), mem or i/o(mem access), cache
level(generic)

Jul 25 11:07:31 node1.local kernel: EDAC k8 MC0: extended error code:
ECC chipkill x4 error

Jul 25 11:07:32 node1.local kernel: EDAC k8 MC0: general bus error:
participating processor(local node response), time-out(no timeout)
memory transaction type(generic read), mem or i/o(mem access), cache
level(generic)

I previously had RedHat 4.0 installed on these nodes and didn't see any
EDAC errors.

I've run memtest86 on the 4 nodes (100 passes) and each of them passed
without error.

Does anyone know if it is possible to disable the EDAC checking and
error reporting?  Primarily because, these errors are logged every
second and in a short time fill up the /var partition.

Thanks for any info,

Mike




[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]