Unsolved
This post is more than 5 years old
214 Posts
0
973
Mirrorview full resync after full DR test
Hi there,
I have a client that is doing a full fail over SRM DR test with Mirrorview/A, however it takes 8 hours to resync with the production side. Is this normal or is there something we could do to improve?
I did notice that the RLP is very small on the DR side, would this have an affect?
Thanks,
Ed
kelleg
4.5K Posts
0
November 16th, 2015 08:00
Please see the following:
https://support.emc.com/docu32906_MirrorView_Knowledgebook:_Releases_30__33_-_A_Detailed_Review.pdf?language=en_US
If the RLP was too small you would see errors - using RLP's also implies that you're probably using Mirrorview/ASYNC.
The settings for a full re-sync can affecting the time - Low, Medium, High. The quality of the link between the two arrays will impact performance. The bandwidth of the link will impact throughput.
The types of LUNs and the speed of the disks will also affect performance.
glen
edhoward
214 Posts
0
November 17th, 2015 03:00
Thanks for the response, I'll feedback and see if that solves the issue.
Regards,
Ed
edhoward
214 Posts
0
November 18th, 2015 02:00
Hi Glen,
They have re-sync set to high and the re-sync is still taking forever. Does doing a fail-over mean that you have to do a full re-sync even if you have only been failed over for a short period?
Ed
kelleg
4.5K Posts
0
November 18th, 2015 14:00
I believe that failover does require a full re-sync - because you never know what might have been written.
glen
dynamox
2 Intern
2 Intern
•
20.4K Posts
0
November 22nd, 2015 06:00
i don't know about MirrorView/A but with MirrorView/S if you fractured your secondary while it was in "Syncronized" state then failback is incremental, if state was "Consistent" than it will require full resync back to original source.