This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Repository Full - Second Repository has more storage available

Hello,

I have an interesting problem.  My Repository is currently full.  I tried to delete some recovery points to free up space, and change retention policies, which helped to free up space, but, my FileServer decided it wants to take a new Base Image, which I don't have enough space to allow it to happen.  I know I could kill all Recovery points from my file server, and let it start again, but I'm hoping to avoid this, as I have a better option.

We have an Outgoing Replication Rapid Recovery server set up that is in the same datacenter (same rack even).  The intention was to put the second server at a second Data Center and have backups replicate...but it never happened.  We have roughly 4 more TB of space on this second server.  I think I want to make this the primary server, but am not really sure how to do it.  I didn't want to do anything drastic, as I'd be curious if there's a really easy way to do it since it's backing up and replicating.  I'm going to try to do some more research, my preliminary research has not found much, and I figured I'd get the question asked, and then I'll post an answer if I find it.

  • It's pretty simple to switch from using one to using the other as the primary.

    1. Pause protection on the source core and ensure replication is caught up
    2. Delete replication from both cores but keep the recovery points. All the machines on the replication core should become recovery points only. 
    3. Protect all the machines on the target core (do not remove them from protection on the original source core becuase that will trigger a new base image for all machines)
    4. Once protected, force a backup and make sure an incremental is taken
    5. Once you confirm that backups are working properly, remove the machines from protection on the old source core
    6. Configure replication from your new source core back to the other core

    I'd recommend doing these steps for one small agent to start with to make sure you understand them and that they work as you expect. Then once you have the steps down, do all of them in bulk.

  • Step 2: I would go to Replication and Delete the Replication Core?  When I delete it, does it then prompt to keep the recovery points?

    Step 3: When you say do this on the target core, do I then go to the target core's weburl, find my VM, and then hit the PRotect VM button up top?  

    Just wanted to clarify, I appreciate your response!  

  • Step 2 - I would start by just deleting one replicated machine. When you select it and click delete, it will launch a confirm dialogue and in that box will be the option to delete recovery points. Don't choose to do that.

    Step 3: Open the target core Web UI, then click the protect VM button at the top. That will launch the protect wizard and you'll need to enter the info for the VM you want to protect. It's exactly the same steps you used when you started protecting the machines when you first set the system up. 

  • Thank you for these answers.  I went in and paused the Outgoing Replication from my current Core to the new Core.  Will I ever turn that back on, or basically just Delete that connection (either now, or once everything is moved?)  I'm going to slowly do this over the weekend, one machine at a time, so I'm not sure if it's ok to pause replication.  We're a small shop, and only backing up like 15 VM's, but a decent chunk of data overall, so some of the transfers will take awhile for sure.  

  • It depends on if you take new backups or not. If you take new backups, you need to make sure those are replicated before you delete the replication and then move the protection of that agent over to the other server. If you don't have all the recovery points replicated when you make the switch, you're going to get a new base image and I assume you want to avoid that.

  • As I do each machine, I'll force a backup as you had recommended.  I just did the first small machine, and it seems like it's all set.  I guess maybe a better way to phrase it (and I think you answered it), Is this:  If I leave the replication from current core turned on and only pause it each time I'm working with a virtual machine (or do I not need to keep Pausing Outbound Replication?), will this work?  I am trying to avoid new base images for most of them, my file server I can't avoid, but there's some other machines that I'd like to not have them take a new base.  My fileserver currently is trying to take a new base image which is causing the space issues.  

  • Yes. That will work. No need to pause it for any machine other than the one you are migrating after it is in sync and you are about to delete it's replication pairing.

  • So do I need to even pause Outbound Replication at all? If so, then I'd imagine I would just turn it back on after a machine is done, and a backup is complete, and let it catch up any existing jobs (or can I just move to the next one).  Sorry for all the questions.

  • I wouldn't pause replication. I would pause the backups for the machine you are gonna move so that no new backups happen while you are doing the migration. Force a backup, let it replicate, then pause the backups, then move the machine by deleting replication and then protecting on the core you want to move it to.