This post is more than 5 years old
172 Posts
0
26871
Incorrect snapshot error messages
I get messages like this from time to time for our scheduled snapshots on a PS6000:
[ID: 4.18] Schedule 4 Hour for volume mesa has fewer than expected snapshots. Expected 8, but the volume only has 0.
But when I check group manager, the proper number of snapshots exist and I have no problem creating a new one. The error is generally kind of random and usually happens for one of two particular LUNs. Also, it usually detects fewer than expected (like 5 instead of 8). This morning, one message listed 5 different LUNs and said all had zero snapshots. I checked each LUN, and all had the proper number of snapshots.
It's waiting for the latest firmware update, but it's not very far out of date. This has been going on for at least a year, so I've mostly been ignoring it since the problem seems to be cosmetic. Is there something else I should check?
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
1
February 24th, 2015 07:00
Hello,
Are these part of a collection? I have seen where a timing error at creation time, that causes this error when they are part of a collection.
Regards,
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 24th, 2015 08:00
You could try that, but I suspect you will end up at same point. What version of EQL FW are you using? If they don't have to be sync'd you can also create individual snapshot schedules instead.
trnc
172 Posts
0
February 24th, 2015 08:00
Yup. Would it do any good to delete the schedule and collection, and then recreate?
trnc
172 Posts
0
February 24th, 2015 08:00
Yup. Would it do any good to delete the collection and schedule, and then recreate?
trnc
172 Posts
0
February 24th, 2015 09:00
I'm at 7.0.9. I'll whack the collection schedules and create individual to see if that helps.
Thanks.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 24th, 2015 09:00
What kind of volumes are these? Are you running the scheduler on the array?
trnc
172 Posts
0
February 24th, 2015 10:00
Yes, it's the built in EQ snapshot scheduler. These are MS terminal server data volumes. I'm backing up nightly to a deduplication device, so the snapshots are mainly for quick recovery in case a server crashes. It was also originally for faster restore than tape for retrieving individual backup files because we didn't have dedup when I created the schedules. Since I can retrieve backup files from dedup a lot faster now, I'll probably change the schedule to only keep a few snapshots.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 24th, 2015 10:00
You might want to consider using the HIT/ME software to increase the quality of your snapshots. As it will call Volume Shadow Service on servers to first quiet the filesystem before taking the EQL snapshot.
The array scheduler only covers what's in cache or on disk. Using HIT/ME software flushes cache to EQL storage first.
Just a suggestion.
Regards,
trnc
172 Posts
0
February 24th, 2015 11:00
Ah, OK, I'll check it out. Thanks for the tip.
Tim
trnc
172 Posts
0
February 24th, 2015 11:00
We're in the process of getting a 6210xs, so I'll be joining it to the 6000. I imagine the 6210 uses the same HIT/ME software, eh? Or is there something different for the hybrid arrays?
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 24th, 2015 11:00
It's all the same regardless of array model #. They all run same FW (when at same revision of course) What size of the drives in the 6000? Are you replacing the 6000?
trnc
172 Posts
0
February 24th, 2015 13:00
2TB drives on the 6000. We still have a couple of years warranty on it, so we'll join the 6210 to it.
Origin3k
4 Operator
4 Operator
•
1.8K Posts
0
February 24th, 2015 13:00
Dont place your 6210xs together with a 6000e into the same pool. The capacity based LB would place most of the data on the 6000e.
Regards,
Joerg
trnc
172 Posts
0
February 24th, 2015 13:00
Thanks, Joerg. We'll put in separate pools.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
February 24th, 2015 13:00
You will need to upgrade the 6000e to 7.1.2 by the way. The 62xx only runs 7.x firmware and can't be downgraded.
I also would put the 6210ss in its own pool. Worst case you could merge them later very quickly. But going other way is a much longer process.
Make sure all the iSCSI interfaces on both the 6000 and 6210 are enabled. And let the RAID initialization complete before moving volumes or data to the 6210.
The only available RAID level for a 6210XS is RAID6-Accelerated.