IBM SVC Snapshot2.0: insufficient mirror bitmap space with SVC 8.6

[pirx]

case 07355397

Running latest Veeam version with IBM storage integration. After storage team updated SVC to version 8.6 a lot of storage snapshots now fail due to insufficient mirror bitmap space. According to IBM this is 'works as designed'. We have 2145-SV2 SVC nodes which can use significantly less memory for this than new SVC models (512MB vs 20GB). FlashCopies are now created on both mirror sides, instead of one as it was before.

Is there a way to fallback to Snapshot1.0 method? Any regkey that can restore the old behavior? The only solution that IBM provided was 'buy new hardware' or create less storage snapshots at the same time...

Code: Select all

[25.07.2024 22:49:12.165]    <01> [00A0]    Info (3)    [SanLunSnapshotCreator] Creating storage snapshot for volume LUN or NFS share: 'VMware_CL21_mirror_WBE_056', '94c21d91-51b7-4eef-b3b1-5da228befad1'
[25.07.2024 22:49:12.165]    <01> [00A0]    Info (3)    [IBM] Creating backup snapshot for volume Id: '16'
[25.07.2024 22:49:12.165]    <01> [00A0]    Info (3)    [IBM] Checking the possibility of using Snapshot2.0 through a new condition
[25.07.2024 22:49:12.337]    <01> [00A0]    Info (3)    [IBM] [SnapshotV2] Create backup snapshot for volume: '16'
[25.07.2024 22:49:12.618]    <01> [00A0]    Info (3)    [IBM] [SnapshotV2] Creating volume snapshot for volume: 'VMware_CL21_mirror_WBE_056(600507680C80802F30000000000418C3)', with name: 'VeeamAUX_VMware_CL21_mirror_WBE_056_2ifRicYHhCyvANwU2oQKW8'
[25.07.2024 22:49:12.618]    <01> [00A0]    Info (3)    [IbmCli] Creating volume Snapshot. Name: 'VeeamAUX_VMware_CL21_mirror_WBE_056_2ifRicYHhCyvANwU2oQKW8', VolumeId: '16'
[25.07.2024 22:49:12.978]    <01> [00A0]   Error (3)    Unable to create volume snapshot 
....
[25.07.2024 22:49:12.978]    <01> [00A0]   Error (3)    CMMVC6350E The command failed because there is insufficient mirror bitmap space.

[david.domask]

Hi pirx,

Thank you for sharing the case number and sorry to hear about the issue; I'm not immediately familiar with a means of changing it, but let's allow the Support Team time to review the situation, as looks like the case was opened only recently. Please include logs from one affected job for Support to review; I understand that you're looking just for a way to reduce the snapshots, but while the behavior is being reviewed, it will be helpful to review the logs to get the full picture.

Thanks!

[pirx]

Sure, I already uploaded them. I also don't want to misuse the R&D forum to escalate this, more an information/warning for others. As I did not find anything about this here yet.

[pirx]

I got information from IBM that there is a regkey that restores v1 functionality and Veeam support provided the key. If the key should not be posted here, please just remove it.

[Moderator: Removed from public forum, please open a Support Case and allow Veeam Support to review the behavior if you suspect you have a similar issue

[david.domask]

Hi pirx,

Great, glad to hear that there was a solution available.