A snapshot is a reference point that preserves the exact folder structure and data content of a data path at a point in time. Snapshots serve as a virtual copy of the data you had at previous points in time and enable you to restore files or directories after deletion or modification.
You can take a snapshot of any directory at any time and you can create snapshot policies to schedule periodic snapshots of specified directories.
Snapshots can be viewed in a hidden read-only directory called .snapshot, which is visible under the directory of which the snapshot is taken.
For example: A snapshot of the path /a/b is created. The snapshot is named MySnapshot_b. The snapshot is visible as a directory called MySnapshot_b under /a/b/.snapshot. The content of /a/b/.snapshot/MySnapshot_b is identical to the content of /a/b at the time of snapshot creation.
You can also choose an option in each view policy to make .snapshot directories accessible in every directory. In subdirectories of protected paths, these .snapshot directories will provide links to any existing snapshots of parent directories, even if there is no protected path on the subdirectory itself. This provides easier access from each directory to snapshots of parent directories.
Snapshots consume space when they hold on to storage that would have otherwise been released. This can happen when files are deleted or existing blocks are overwritten with new data. In these cases, the snapshot holds onto the original data.
When you take a snapshot on a given path, the snapshot preserves the data under that path only. Its consumption depends only on the delta between the current data under that path at any given time and the data under that same path at the point in time when the snapshot was taken.
VMS provides the following metrics to estimate the capacity consumption of snapshots:
Per snapshot, you can see:
Aggregated Usage. This field shows an estimate of the amount of usable capacity that could be reclaimed by deleting the snapshot assuming that all snapshots on the same protected path that are older than the snapshot were already deleted.
Note that if you have snapshots on multiple protected paths that are holding some of the same data, such as two snapshots on nested protected paths taken close in time, that data will not be counted as reclaimable by this estimation for any of the snapshots.
Consider an example: Suppose there is a protected path on
/A/B. A snapshot, which we will call 'X' was taken on the protected path when there was 1GB of data residing under the path. Suppose snapshot X is the only snapshot on
/A/B. Snapshot X 's Aggregated Usage will be 1GB. Now suppose that there is also a protected path on
/Awhich also has one snapshot ('Y') which was also taken at the same point in time as snapshot X. Snapshot Y protects the same 1GB of data from being deleted as snapshot X. In this case, deleting snapshot X would not reclaim any of the 1GB and therefore snapshot X's Aggregated Usage will be 0GB. Aggregated Usage for snapshot Y will also be 0GB unless
A/is holding onto additional data, such as data under
Unique Usage. This field shows an estimate of the amount of usable capacity that could be reclaimed by deleting the snapshot without deleting any other snapshot, including older snapshots on the same protected path.
To see this in the VAST Web UI, browse to the Snapshots page (Data Protection > Snapshots). Each of the fields is a column heading.
If either of the columns is not displayed, click the Choose Columns to Display button ( ) and select the column names that you want to display.
Per protected path, you can see:
Aggregated Usage. This field shows an estimate of the amount of usable capacity that could be reclaimed by deleting all snapshots on the protected path.
This estimation takes into account any nested protected paths that hold common data, because data held by another protected path's snapshots would not be removed even if all snapshots on the protected path were removed.
For the capacity consumption of all snapshots on the cluster combined, look at the Auxiliary used physical or usable capacity estimate for the cluster. This is the amount of usable SSD space in use by (a) snapshots and (b) any data that may be pending deletion at the current time. This figure usually gives a close indication of snapshot capacity usage. However, at times when a large amount of data was recently deleted, wiping that data from physical storage may taken some time and temporarily enlarge this figure.
In the VAST Web UI you can view the auxiliary used metric in the Capacity card in the Dashboard under the Usable heading. If Show usable capacity is disabled in VMS settings, then you can see a similar and probably larger value under the Physical heading.
Both usable capacity and physical capacity metrics report physical drive capacity. The difference is that physical capacity includes capacity used for data protection and background activities, while total usable capacity is the amount of physical capacity available after subtracting the amount of capacity required as overhead for data protection and background activities.
Since snapshots only indirectly consume storage (by holding onto storage that would otherwise be released), you cannot directly control the storage consumed by snapshots. However, you can effectively limit the amount of storage consumed when you set up a snapshot policy for each data path; you choose how frequently a new snapshot is taken and how many snapshots to retain for that path.
No. When you take a snapshot of a directory, pre-existing snapshots of subdirectories are not included in the data preserved by the snapshot.
Snapshots are part of the logical realm. Managers with permission for the logical realm can manage snapshots. Managers who have object-level permission to access a specific export can also manage snapshots for that export.
If a snapshot is flagged indestructible, it cannot be deleted or modified unless the cluster's indestructibility mechanism is unlocked by specially authorized personnel.
The following capabilities are available:
Capabilities available through VMS:
Fast Restore, in which you restore a given path from a snapshot. The snapshot can be from a local protected path or from a protected path that is replicated to the local cluster from a peer. A restore can be aborted during the procedure until the last step.
Global snapshot cloning, in which you create an instantly writable clone of a snapshot. The snapshot can reside on the same cluster or on a different cluster that has a replication peer configuration with the cluster where the snapshot resides.
Restore capability for client users: Client users can restore files and directories from NFS and SMB clients.
Yes, you can flag a snapshot as indestructible. It cannot be deleted and its expiration time cannot be shortened without a cluster-wide unlocking of the indestructibility mechanism. For information about indestructibility, see Keeping Indestructible Backups.
Article is closed for comments.