We have 2 volumes which are limited to keep a maximum of 2 snapshots, however these go up to 8 or 9 snapshots. We can see the 2 scheduled snapshots but also the other snapshots taken at various times throughout the day. The snapshot times do not match with the scheduled times in ASM on the hosts.
The snapshot policy is not set to 'Borrow snapshot space as needed' and set to 'Delete oldest snapshot'.
We have upgraded ASM on the hosts to Version 22.214.171.12428, and the storage array firmware to version 8.1.10 (R436828) hoping this would resolve the issue but it hasn't.
Any help would be appreciated?
That's great news, thanks for the update.
re:1. Group Manager based snapshots can only capture the data that's on the array storage already or in the cache of the storage. Any pending writes on the host wouldn't be included. When you use ASM/ME the Volume Shadow Service is called first. This pauses I/O and flushes out cache, then a snapshot is created. Making it more consistent. A GUI based snapshot is said to be "crash" consistent. At least as consistent as would be if the server crashed or lost power. Using ASM/ME it's said to be "application" consistent. Also ASM/ME integrates with MS SQL, Exchange and SharePoint to provide more consistency and additional features. I.e. recover a single Exchange Mailbox entirely via the ASM GUI.
Re: 2. It sounds like there is a virtual disk on local storage not on the SAN. So if host A has a piece of the VM and you move to host B. Host B can't quiet the filesystem on Host A.
You probably want to open a support case for that. So they can review the logs and take a look.
Social Media and Community Professional#IWork4DellGet Support on Twitter - @dellcarespro
The scheduler will only trim snapshots taken by that schedule. So for example if that's a once / day schedule and you have another running somewhere doing every 2 hours, the 1/day schedule won't trim any of the 2 hr snapshots. Only if you run out of snapshot reserve would one schedule prune the oldest snapshots to free up space.
That makes sense, the scheduled snapshots never increase more than 2. The volume is used as a cluster shared volume with multiple Hyper-v VM's. Is there anyway to find out where these other snapshots come from? They don't have a schedule name next to them.
I would first check the Group Schedules to see if someone created a a local schedule. If it com,es externally I would expect a login event at the snapshot creation time. It will show you what IP address is came from.
The other snapshots had the description 'Created be the EqualLogic VSS HW Provider from %hostname%'
I disabled all the schedules on the ASM on the hosts and the snapshots did not re-appear in the EqualLogic group manager. The ASM schedule has 14 VM's but only 6-7 snapshots showed in the EL group manager. Not sure how that all works but that seemed to be the cause.
On a side note,
1. Whats the difference between group manager snapshots, ASM - schedules for VM's,smart copies and applications?
2.On the ASM most of the scheduled VM's have yellow triangles with an error saying 'A component or volume referenced by this schedule is missing on this computer. I can see the the VM's which are ok are on the correct host, but VM's failover to different hosts. Do these schedules have to be reconfigured or do we have to move the VM's back to their original hosts?
Thanks for you help, I will call Dell support.
You are very welcome. I"m glad I could help.