Unsolved
This post is more than 5 years old
44 Posts
0
1017
Question regarding recoverpoint image test & failover
Hello experts,
Can i ask 3 questions regarding recoverpoint.
1、During test the copy at the DR site(logged access mode with VPLEX splitter), filesystems are mounted on the DR server, if i make some change on the mounted filesystem at DR side, how can i reserve the change and replicate it back to the Prod copy ? Normally, if when close the test, all the change on the DR server will be lost(rollback).
2、Starting RP 4.0, the failover procedure was divided into two steps(i.e. image test and actual failover), the wizard will ask you to umount the filesystems at the DR server after image test, then go to actual failover. My questions is, if i make some change during the image test, then failover the CG, will the change be reserved ? Or it will be lost just like normal image test ?
3、Prior to RP 4.0, we can do failover during image access directly, before failover it will not ask you to umount the filesystems at the DR server, will the change made during image access be reserved ? if Not, will the DR server has some filesystem problem because of the unexpected data change ? Suppose i don't umount the filesystems before actual failover.
Thank you in advance.
Dong_JH
44 Posts
0
August 25th, 2017 02:00
Anybody here ?
forshr
1.1K Posts
0
August 25th, 2017 05:00
1. By failing over by invoking a failover via CLI or Rest API.
2/3. At 4.0 the flow changed (more closely integrated) as a safeguard which is why you would need to use the failover action via the CLI or Rest API command if you have invoked image access, make changes and want to keep and replicate this data back to the original Production.
Regards,
Rich
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
August 29th, 2017 12:00
Hi,
For question 1: Another workaround would be to enable direct access on the copy lun and make the changes. There are a few caveats if you are considering this approach. First, you will lose all existing journal point in time snaps, so no recovery options once you enable direct access. And secondly, you will have to go through a short initialization if the image is not suitable and you want to resume replication in the normal direction (the workaround for this caveat, would be to check the image in image access mode before enabling direct access). Thirdly, once direct access is enabled, in order to replicate in the opposite direction, you will have to disable and delete the existing CG and recreate it with the copy lun as the new source.
Thanks,
Tom Pinto.