From my knowledge. snapvault is actually a snapshot but not in the box or controller, but outside the box. so you need 2 n series to do snapvault. the diffrence between snapvault and snapmirror is almost the same both using snapshot technology, but snapmirror is replication from primary controller and secondary controller and snap mirror use the snap update to compare the snapshot between secondary and the primaryif there are any changes at the block then snapmirror will update the snapshot at the secondary site. and snapvault is actually the same with snapshot but the location where the snapshot was taken is at the secondary controller.
i hope it answer your question, sorry for my messy english, and just CMIIW
snapmirror is a DR application whereas snapvault is a backup application.
When you setup snapmirror schedule, snapmirror creates a (incremental) snapshot on the primary as part of the update and mirrors it to the secondary. This incremental snapshot becomes the base snapshot for the next update. The data being updated is always the delta between the base and the incremental. This way, snapmirror ends up replicating only the data that is changed beteween base and incremental snapshots. At the end of every update, the old base snapshot is removed on both primary and secondary and new incremental snapshot is marked as the base of next update. So, alwayes a snapmirror relationship retains one snapshot as the base of next update.
SnapVault uses the same engine as qtree snapmirror to replicate snapshot data. But there are some key differences.
- You need an independent snapshot schedule on the primary. snapshots are not takes as part of snapvailt update.
- snapvault snapshot schedules are per volume unlke snapmirror schedules which were per relationship
When the snpavault schedule on the secondary kicks in, it initiates an update to *every* qtree on the secondary volume. The primary of all those qtrees can come from different volumes/storage controllers etc. Once all qtrees are updated, snapvault will create one archive snapshot for the volume and keeps this snapshot till the retention count specifed in the schedule is reached. Hence snapvault will keep retention count number of snapshots in the volume - unlike snapmirror which keeps only one. This way, with snapvault, you can theoretically have more than 256 number of relationships per volume.
If the question is whether you can use the same volume both as snapmirror and snap vault destination at the same time - no, you can not. You can use different qtrees on the same volume for QSM and SV, but this won't achieve space savings you mean.