We are haing an issue with how much data is getting snapmirrored to our DR site and until we figure it out we will never catch up on our lagtime. Can we just delete a daily snapshot folder so that that sill not get transferred or is that a bad thing. We need to stop all transfers remove teh last couple backups and start fresh. Is there an easier way to get this accomplished? Any help is appreciated. Thanks
*We are haivng an issue with how much data is getting snapmirrored to our DR site*
This seems to be something that will not get solved by my answer below. You need to make sure that the WAN bandwidth and the amount of data you need to transfer within the time you have are matching. Sometimes one needs to add bandwidth to solve this, or change the volume layout or use compression with SnapMirror... Whatever fits your needs best.
You can delete some of the daily,hourly,weekly snapshots to optimise the transfer of changed blocks, but his will probably not make up a lot of blocks.
Use the command
filer> snap delta "volname"
to find out how much changed data you have per snapshot and then find a way to minimize the data change within this volume, or see first part of the answer.
Thanks for the reply. I just wanted to clarify a couple things i may have left out. We run snapmirror and snapvault jobs nightly. The snapmirror job only runs one nightly job for a 7 day period. this is strictly for DR purposes. That snapmirror jobs are running with compression enabled. We have 3 datastores that get snapmirrored and they were all about 10 to 12GB in size which was ok but then all of the sudden 2 of teh datastores went up to 25 to 30GB. now the lag times are too long since the jobs can't finish in a timely manner. It is only the VMare side of the backup that i am concerned with. We run snapvault for our CIFS data and snapmirror for all the VMware data. As for what i can delete or not, my .SNAPSHOT folder on the vmware side only has 8 snapshot folders. I was curious if i was able to stop the job, remove one or two of the folders, then restart teh job without causing any issues. I am trying to figure out why the size has changed but until then i need to find a way to get back on a schedule. I have tweaked all the data so that all the swap locations are on a differnt datastore that doesn't get replicated. Anyhow, that is what i'm looking to do and not sure if i'm able to do it without issues. Thanks again for teh reply.
Well, i'm not really sure what you meant by remove it from SMVI. I know what SMVI is but not sure how those certain folders get removed. Are you referring to removing teh backup job and recreating it?? Got one more question concerning this issue. From the attachment below can you tell me if both np_Datastore0 and Lutz_Datastore0 are both being replicated? I have all the SWAp Volumes on another datastore and they are nos supposed to be replicated to the DR site. Having both of those boxes checked in the backup seems to me that both are geting done. Any thoughts??
In SMVI there is a list of available backups (in the restore section). There you have the option to solect a backup and delete it. This will also delete the correxponding snapshot on the storage system. You do not need to delete the job.
There seem to be three volumes being replicated:
The np_datastore0 is not being replicated, only local snapshots are being created from SMVI Backups.