Table of ContentsView in Frames

Limit on the number of concurrent SnapVault targets

Starting with Data ONTAP 7.3, the maximum number of concurrent SnapVault transfers of individual qtrees on a storage system is increased. However, the maximum number of concurrent SnapVault targets has not changed.

Before Data ONTAP 7.3, the maximum number of concurrent SnapVault targets supported by a system was equal to the maximum number of concurrent SnapVault transfers possible for the system.

A SnapVault target controls the creation of a scheduled SnapVault Snapshot copy on a SnapVault destination volume. For each SnapVault destination volume that has qtrees being updated, there is a SnapVault target.

The maximum number of concurrent SnapVault targets for each platform is described in the following table. At any point in time, only the number of volumes listed can have their qtrees updated concurrently. If the number of SnapVault targets exceeds the limit, the excess SnapVault targets are queued and executed after the active SnapVault targets complete their backups.
Model Recommended maximum number of concurrent SnapVault targets
FAS2220 128
FAS2240 128
3140 128
3160 256
3170 256
6040 192
6080 256
3210 128
3220 256
3240 256
3250 256
3270 256
6210 256
6220 256
6240 256
6250 256
6280 256
6290 256
FAS8020 256
FAS8040 256
FAS8060 256
CBvM100, FDvM100 32
Note: These maximum numbers apply only to SnapVault targets, and therefore to SnapVault qtrees. There is no restriction on the number of volumes that can be updated concurrently for SnapMirror qtrees.

Although there is a maximum number of concurrent SnapVault targets, you can configure SnapVault relationships in as many volumes as required. However, at any point in time, only the qtrees in the limited number of volumes can be updated.

If the SnapVault backup throughput is limited by the bandwidth or latency of individual links between multiple primary systems and the secondary system, you can use the enhanced number of concurrent transfer limits available in Data ONTAP 7.3 and later releases to get higher overall throughput. You can use the higher limits by spreading the qtrees in the specified number of volumes.