I am very new to NetApp Technologies and learning things.
In Our Current Scenario -
NetApp filer is configured as backup destination for one of our Enterprise Backup Tool that we use for primary snapshots backups ( wherein VSS writers are used for snapshots ). Whenever backup retention reaches it will delete all old backups and its relevant snapshots to free space on filer accordingly. However , NetApp filer is taking sometime to show freed up space.
Is it true "NetApp controller will take sometime at the volume level to get the blocks reclaimed but the messaging either on command prompt or via System Manager should show freed up space immediately"
When you mention that it is "taking some time to show freed up space", are you referring to System Manager? System Manager always takes a few seconds to refresh. For large deletes I have seen the space gradually show as available over a bit of time (a few minutes maybe) but nothing that would cause any operational issues. The CLI commands are typically more accurate on an instantaneous basis but the delta is not usually that great.
Is there an operational issue that you are trying to solve?
It is true that when data is deleted, it will take some time before any method of capacity reporting will accurately reflect the recent deletion of data. I have seen this in both deleting a volume and also deleting snapshots within a volume. Think of it as a soft delete or a "quick format" kind of deletion in that the pointers are removed to the data blocks but in the background the blocks are slowly being wiped or permenantly deleted. Of course this will be more obvious if you are removing a large amount of data rather than a small amount but if you run capacity cmdlets consistently you should notice the free space in either case slowly increase as the blocks are freed up in WAFL. I have also confirmed this is the case with netapp support because they commented on a background process that is running doing the actual removal of data, said process cannot be stopped once a delete command has been issued in the case of a volume or a snapshot. hopefully that helps you understand WAFL's typical behavior as it relates to data removal.
Hi can you elaborate on this? Say I have an aggregate snapshot that's taking up 100 GB, and I delete it. Now, is that space immediately returned to the aggregate as free space or does a background process run and zero the blocks or the like? I'd imagine there must be some kind of zeroing going on as if those blocks got assigned as a new volume and the data was still in there it might be a security risk. Can this be seen using one of the WAFL privileged command? My concern is that I'm getting alerts from DFM about aggregate free space being exhausted so I deleted some flexclones/snapshots in the aggregate etc. but the free space does not seem to diminished until 30 mins pass. I'm wondering if this is due to this block reclamation? This is causing me issues as when the space in the aggregate goes down to 0, the LUNS in my flexclones are brought offline.