Unsolved
This post is more than 5 years old
5 Practitioner
•
274.2K Posts
0
2174
Recoverpoint for Virtual Machines with VMware HA
Hi!
I have a customer who is currently running a VPLEX Metro environment with two VNX 5400 systems in two sites.
They are now looking for a cost-efficient way to deploy a third site for disaster recovery purposes.
The idea is to use three existing stand alone servers with ScaleIO (about 50 drives in total) and implement RecoverPoint for Virtual Machines, connecting the third site to one of the existing sites.
I am finding it hard to find any documentation to wether this is possible or not. Can RecoverPoint for Virtual Machines be used on VM's that currently exist on top of a VPLEX Metro environment? I could only find that Fault Tolerance is unsupported with RP4VM, nothing about HA support.
Any help is much appreciated!
Thank you!
forshr
1.1K Posts
1
June 1st, 2015 08:00
The proposed solution, if I'm understanding correctly, is combining traditional RP with RP4VMs with RP4VMs at the DR site? If this is the case, then this is currently not possible to combine the two code streams.
However, if you are proposing to perform simultaneous replication of a source VM to a 3rd site by layering RP4VMs over the top of the current VPLEX Metro with traditional RP solution, then this would be subject to an RPQ approval. It will be up to the customer to get the VPLEX consistency group rules to match where splitting is happening (i.e. determine which site has the ESXi host doing the splitting and set the VPLEX site preference rule to match).
Regards,
Rich
forshr
1.1K Posts
1
June 1st, 2015 10:00
So the second part of my answers refers to this type of configuration. RP4VMs is storage agnostic and therefore does not know that the VPLEX Metro is there BUT at this time this configuration does require an RPQ. Ultimately this comes down to the customer understanding and appreciating the impact of initiating any RP4VMs recovery activities on the VPLEX Metro or vice-versa.
Anonymous
5 Practitioner
5 Practitioner
•
274.2K Posts
0
June 1st, 2015 10:00
Thanks, Rich!
Just to clarify; we're proposing going RP4VM on all three sites. Two of them are existing sites with VPLEX Metro.
Is this possible with RPQ?
Kind Regards
forshr
1.1K Posts
0
June 1st, 2015 11:00
I should add that you will have to wait for the RP4VMs version 4.3 planned to GA around mid-July to allow for the 3 sites with 3 vRP clusters.
padukone
35 Posts
0
January 15th, 2016 08:00
Hello,
We have a similar environment. ESX hosts are in a stretched cluster across sites using storage from VPLEX metro. We didn't want to go down the road of RPQ so we are trying to create 2 RP clusters, one for each site. RP-cluster1 will protect all the hosts at Site A and RP-cluster2 at Site B. However, we still need to use a distributed vplex volume (one that is seem by all the hosts in the strected ESX cluster) as repository datastore, because the deployment manager only recognizes datastores seen by all the hosts.
My question: Does RP4VM support using a distributed vplex volume as a repository datastore? I ask this because we tried using distributed volume and the deployment manager fails with an Exception :Failed executing CreateAndAttachRPDeviceAction.
Idan
675 Posts
0
January 15th, 2016 09:00
Hi there,
Altough technically possible, placing journals and repository on distributed volumes (or any other vMSC cluster) requires a RPQ.
The error you're getting seems unrelated to the fact that the proposed DS for the repository is a DR1.
Hope that helps,
Idan Kentor
RecoverPoint Corporate Systems Engineering
Twitter: @IdanKentor