Dell EqualLogic PS4210X EqualLogic Group Manager Administrator s Guide PS Seri - Page 313
Automatic Snapshot Creation, About Synchronous Replication Switches and Failovers
View all Dell EqualLogic PS4210X manuals
Add to My Manuals
Save this manual to your list of manuals |
Page 313 highlights
NOTE: If data reduction has been enabled on the volume, snapshot reserve is permanently disabled. When you create a snapshot of a volume for which synchronous replication (SyncRep) is enabled, the snapshot resides in the pool that is the SyncActive pool at the time the snapshot is created. If the SyncActive pool is switched, existing snapshots remain in the pool in which they were created, and subsequent snapshots reside in the new SyncActive pool. You can perform the same operations on any of the volume's snapshots, regardless of whether they reside in the SyncAlternate or SyncActive pool, with one exception: Only snapshots residing in the SyncActive pool can be restored. To restore a snapshot in the SyncAlternate pool, you must first switch pools. For example, assume a synchronous replication volume named Volume1 has a SyncActive pool named PoolA and a SyncAlternate pool named PoolB. The group administrator takes 5 snapshots of the volume. All of the snapshots reside in PoolA. Later, the group administrator switches pools, and creates 3 snapshots. These new snapshots reside in PoolB. Now, 5 snapshots are in PoolA and 3 snapshots are in PoolB. The group administrator subsequently switches back to the original configuration and creates 2 additional snapshots that reside in PoolA. Now, 7 snapshots are in PoolA and 3 snapshots are in PoolB. Automatic Snapshot Creation The group creates snapshots of synchronous replication volumes in the following circumstances: • Anytime an administrator performs a pool switch, failover, or disconnect • Whenever the volume's state changes to paused or out of sync • Just before the group begins writing tracked changes to a the SyncAlternate pool in an out-of-sync volume These snapshots are called protected system snapshots and failback system snapshots, and they are visible in the GUI and CLI. System snapshots reflect the volume's data at the point in time when the operation was performed or the volume's status changed. Depending on their type, system snapshots can reside in either the SyncActive or the SyncAlternate pool. If the SyncActive pool snapshot reserve does not have enough space, the group does not create a snapshot. Not creating a snapshot does not have any impact on the operation or status change, or on the volume's availability. Snapshots created during a pool switch are deleted after the switch has completed successfully and the volume's status returns to in sync. If the SyncAlternate pool snapshot reserve does not have enough space for a system snapshot, the group cannot put an out-of-sync volume back in sync. To ensure that a volume can be put back in sync, Dell recommends that snapshot borrowing be enabled in groups where synchronous replication is in use, and that the snapshot reserve for synchronous replication volumes be set to 100 percent or more of the volume's size. About Synchronous Replication Switches and Failovers You can change iSCSI initiator access from the SyncActive volume to the SyncAlternate volume using either switches or failovers. NOTE: Although no changes to the volume's iSCSI initiator configuration are required, the change might disrupt host connectivity to the volume. In some cases, you might have to restart host applications, iSCSI initiators, or operating systems for hosts to reconnect to the volume. Consequently, Dell recommends that you perform synchronous replication (SyncRep) switches and failovers during periods of downtime or minimal activity. SyncRep Switches A pool switch for a synchronous replication volume swaps the roles of the SyncActive and SyncAlternate volumes. As part of the operation, the SyncActive volume becomes the SyncAlternate volume and the SyncAlternate volume becomes the SyncActive volume. When a pool switch takes place, iSCSI initiators are temporarily logged out of the volume. They can log back in again after the switch is complete. Depending on the initiator configuration, the initiators might automatically attempt to reconnect. After the failover or switch completes, the volume is automatically set online and synchronous replication resumes. About Synchronous Replication 313