Unsolved
This post is more than 5 years old
8 Posts
1
5620
Error email received at end of scheduled auto-tier window
We have a VNX 5300 with a single storage pool that is configured with a scheduled auto-tier operation each night. Since upgrading to 5.32.0.5.015, we receive an error email each night with the following contents. The email is sent at the conclusion of the auto-tier schedule window (verified by changing the windows size from 4 hours to 2 hours, the email time changes as well).
Time Stamp 02/06/13 05:00:01 (GMT)
Event Number 712d841a
Severity Error
Host SPB
Storage Array APM00112503889 SP N/A
Device N/A
Description Internal Information Only. Could not complete operation Relocate 0xB00037E7A Cancelled CBFS status 0x26 because 0xe12d8426.
00000400 03002c00 d3040000 1a842de1 1a842de1 00000000 00000000 00000000 00000000 00000000 712d841a
We started working with EMC Live Chat support earlier today in hopes of resolving the issue, but I don't beleive we have a valid answer yet. The technician has referenced a couple different Powerlink documents, but neither seems to be correct for the 0xe12d8426 extended code we're receiving.
Has anyone else experienced this error? How to resolve it?
sphilp1
8 Posts
0
February 6th, 2013 21:00
The first document that they pointed to was http://csgateway.emc.com/primus.asp?id=emc286486 the other was http://csgateway.emc.com/primus.asp?id=emc291220
Neither of these documents matched the extended code we receive. The first document looks to refer to pre-5.32 FLARE code.
SushantGulati
55 Posts
0
February 6th, 2013 21:00
Can you please provide the link to the Powerlink articles? I think this is a known issue and I remember facing it in September when we did an OE upgrade on VNX.
jim_blackburn
2 Posts
0
March 24th, 2015 07:00
I recently encountered event number 712d841a on a VNX5800 running R33, with logs mentioning reason code 0xe12d8426. EMC support has requested SPCollect files and dump files, suspecting "...the problem is related to a known issue with FAST VP that is occurring on some VNXs with R33 code."
@sphilp, would you happen to have received a workaround or fix for this defect from EMC?
kelleg
4.5K Posts
0
March 24th, 2015 09:00
Please see the following KB article - this applies to VNX running Flare 32 (not the VNX2):
https://support.emc.com/kb/91463
I'd also strongly recommend that you upgrade the Flare version - you're running at 32.015 - that's really old and contains known issues that have been fixed in the latest release (32.216).
glen
sphilp1
8 Posts
0
March 24th, 2015 09:00
Our workaround ended up being extending the scheduled relocation window for the pool.
My recollection was that the event is reported when the end of the relocation window is reached but there are still pending relocations. You'd think that would be a perfectly normal event and wouldn't merit an email, but here we are...
We still occasionally receive the event, but that has cut down the number of instances significantly (usually just when we've done a large data change on the pool).
Hope that helps!
sphilp1
8 Posts
0
March 24th, 2015 09:00
@glen The initial question was asked 2 years ago.
And it's pretty sad that it's still relevant today...
jim_blackburn
2 Posts
0
March 24th, 2015 10:00
@sphilip, maybe there is a difference in the problems we observed. The problem I've observed includes a large volume of slice relocation failures that occur all the time; they are not limited to the end of a scheduled relocation period.
kelleg
4.5K Posts
0
March 25th, 2015 14:00
The problem with Auto-Tiering ending while there are still slices to relocate would not change over time.
The other part of this is that a slice can contain blocks of data, but the slice has free space, but in total, the contiguous free space in a single slice may not be sufficient to hold all the blocks that need to be relocated.
Similar to the issue that we used to see on Windows FAT when the cluster size resulted in wasted space. I believe engineering may be working on a way to reduce this.
glen