Scenario #1
In a 2-Node, non-clustering configuration (1×1) replicating a 100TB volume between Source server and Target server over a WAN connection, the network goes down for twenty minutes.
Question
In this scenario, what would happen to the Mirror State with DataKeeper Standard Edition?
Answer
After a couple of minutes, the Source server will detect that the network is down and the mirror will go from the MIRRORING state to the PAUSED state.
Question
Does DataKeeper continue to track changes on the Source server?
Answer
Yes. The Bitmap (# of Dirty Sectors) will continue to be updated on the Source server while the mirror is in the PAUSED state.
Question
Once the network is resumed, will a partial resync to the Target server occur?
Answer
Yes. The mirror will go to the RESYNC state and remain there until all dirty sectors are written to the Target server. It will be a partial resync.
Scenario #2
In a 2-Node, non-clustering configuration (1×1) replicating a 100TB volume between Source and Target over a WAN connection, the network goes down for twelve hours. The Source server is rebooted while the network is down.
Question
In this scenario, what would happen to the status of the Source server in DataKeeper Standard Edition?
Answer
The Bitmap on the Source server is persistent (on disk), so it will not be affected by a Source reboot. Only a partial resync is needed if the Source server is rebooted. The Target server will report that it is in the MIRRORING state until it is reconnected to the Source server. Then it will go to the RESYNC state while the resync is proceeding.
Post your comment on this topic.