This post is more than 5 years old
36 Posts
0
4403
April 13th, 2015 06:00
Question about - pausing the replication for RPA
Hello team,
Next week, we have an maintenance window for our NETWORK devices for the Core switches.
And we have the 4 node RPA cluster replicating production data < --- > to other DC over WAN link.
So, because of Network maintenance activity, as of safer side, we are thinking of putting the RPA DCG's into PAUSED mode.
So, I want to know that if we put DCG's on paused mode for a longer time say 30 mins or 1 hr. Will it cause the FULL SWEEP or just an incremental sync. up.
Please comment and guide !!
No Events found!
forshr
2 Intern
•
1.1K Posts
0
April 14th, 2015 02:00
As you will be marking metadata in the local copy journal you cannot mark too many changes. However, a lot of changes may constitute what is termed as a long resync and looks like a full sweep (but it isn't) as the total of accumulated data to be replication is deemed to be too big for the journal and is written directly to the replica and as a consequence any previous journal history is lost.
forshr
2 Intern
•
1.1K Posts
1
April 13th, 2015 07:00
Hi Anil,
Pausing any CG, normal or distributed, invokes marking mode to the Prod copy journal. When you resume, a "short init" will follow and not a full sweep.
Regards,
Rich Forshaw
Consultant Corporate Systems Engineer - RecoverPoint & VPLEX (EMEA)
Data Protection and Availability Solutions
Anil2sharma
36 Posts
0
April 13th, 2015 23:00
Hello Forshr,
Thanks for helping in right direction..
I have just one more doubt, suppose in case the prod. copy journal mark too many changes to large time of pause duration or due to too many changes from the host side.
So in that scenario will it go for FULL SWEEP.
And if yes, is there way to calculate or monitor this situation. so that we can avoid the full sweep ...
either by adding more Journal volumes both side
Or by resuming the replication of CG before the FULL sweep condition...