This post is more than 5 years old
32 Posts
0
2144
How to provide host access on replicated LUN using recover point SE
We have recently integrated RP SE in our environment and I would like to know how to provide host access on replicated lun on DR site. I understand that this may have been discussed multiple times but I was unable to find any recent discussion where this question was answered.
I have VNX 5500 and RPA cluster at both sites. I have single CG and in which there are multiple group named 1,2,3,4 and out of them the LUNS which are part of group 1 needs to have host access at DR site. Luns which are part of group 1 are already part of Storage group at DR site and host who need access to those LUNS are also part of same storage group,
I wanted to know how to provide those host at DR site access to those replicated LUN so that they can check the integrity of replicated LUN.
Thanks in advance
forshr
1.1K Posts
0
January 26th, 2014 01:00
Hi Ivan,
There are two steps you need to perform to ensure your DR host can see the replica volumes.
Firstly, make sure the DR host visibility of the replica volumes in its Storage Group. The CG creation wizard will provide access to the RPAs by automatically adding the volumes to the RP Storage Group.
Secondly, you need to use the virtual or logged image access feature in order to invoke R/W access for the replica volumes.
Regards,
Richard
ivan_skinner
32 Posts
0
January 26th, 2014 10:00
Hi Forshr thank you for your reply, I have another query.
1)Once host has performed there testing and we need to revoke the R/W access in that case simply we have to disable the image access and start the replication or is there any other process that needs to be followed ?
2) What will happen to the DR journal logs will that be wiped out once we start the replication?
Thanks
forshr
1.1K Posts
2
January 26th, 2014 12:00
Hi Ivan,
1) Yes. Disabling image access for the most part will simply allow target side distribution to catch up. One of the USP's of RP is the ability to easily test the integrity of replicated data via image access whilsit at the same time continue to allow replication albeit without fully distributing it at the target side.
2) Replication should not be stopped/paused although this is dependent on the size of the journal versus and write rate versus the time spent in image access log. So the journal history shoudl be maintained and new snapshots which were partially written to the journal during the period of image access (actually written to the 'Do' queue of the journal - you can visibily see this by loking at the journal tab of the copy and seeing journal lag increase) are fully distributed to the replica.
Regards,
Richard