Hi all, having studying both CDP and VDRO a bit, I know the former is focusing on data protection with continuous restore points; while the latter focus on DR automation likes Re-IP and VM bootup priority setting.
- can I conclude both CDP and VDRO are required for a complete DR solution?
- any use case only CDP is required or just VDRO is required?
- if our customer ask about CDP restore point keeping 24hrs is long enough, could I say those VMs are still being protected by backup?
1. No you don't need both applications for a DR solution as CDP is it's own DR solution.
2. Use case for CDP would be needing near instant recovery very low RPO and RTO.
3. Yes as CDP is near line replication so also in theory a backup.
https://helpcenter.veeam.com/docs/backup/vsphere/cdp_replication.html?ver=120
Hey
VBR with CDP replication will allow you to setup replication between source and a target. It will also let you create failover plans to say when triggered start up these items and do these tasks (like reIP) in a particular order. To make that start though requires either manual intervention on your part (right click, failover now) or some automation built on your side to say if this event happens then begin failover plan. If you don’t want to create that automation yourself that is where VDRO is going to come into play.
I will add that VDRO currently doesn’t support VMware Cloud Director or Veeam Cloud Connect so if that is where you are replicating to (as you should) then VDRO is off the table for now.
Technically, neither are required. VDRO automates failover and performs failover testing. You can failover within VBR as well, but it’s not as automated, so really it just orchestrates testing and failover.
CDP provides very fast replication with as little as 15 seconds RPO. But that said, you can also failover using snapshot-based replication. VDRO can orchestrate either replication methodology. Personally, I use CDP replication for data that I want/need to have as little delay as possible (SQL servers in my case) but the rest of my servers are replicated via snapshot to the tune of every 4 hours or 12 hours or 24 hours or whatever. It’s really about which architecture fits your needs. It used to be that my critical VM’s were replicating every 4 hours, and everything else every 24 hours, but since I have the capability in both features and bandwidth, I now replicate via CDP with a RPO of something like 30 or 60 seconds and the snapshot replication occurs I think every 12 hours.
Thanks guys give me so much insight of thinking about the solution
Can I say, if my solution somehow catered the replicaiton funcitons, e.g. using MetroCluster at array level, using vSphere Replication at ESXi level… generally it’s DR solution already
Depedns on case, say if many VMs are need to be autoamted, I add VDRO
If continuous restore points required, I add CDP
Thanks guys give me so much insight of thinking about the solution
Can I say, if my solution somehow catered the replicaiton funcitons, e.g. using MetroCluster at array level, using vSphere Replication at ESXi level… generally it’s DR solution already
Depedns on case, say if many VMs are need to be autoamted, I add VDRO
If continuous restore points required, I add CDP
Yes all those scenarios work for your case and examples.
Enter your E-mail address. We'll send you an e-mail with instructions to reset your password.
