Options for small Rapid Recovery 6.* deployment. Server/Storage

Hello,  I am dealing with an aging server hosting Rapid Recovery running Server 2012 and sending backups to a Drobo B1200i over iSCSI.

I am only backing up 5 servers, 2 of which are lightly used file servers. (One of those will be going away soon)

My initial thought was to just buy a decent workstation to host Rapid Recovery Core, and then some type of new storage to send backups to. (our footprint for backups is only about 7 TB)

I currently replicate to a workstation at another physical location across town via a 100 Mb/s fiber connection.

My dilemma right now is choosing some decent, but low cost storage and whether or not it should be direct attached, or also iSCSI. 

Appreciate any thoughts on the matter.

  • Leaning towards a Synology DiskStation DS920+ with 3 WD RED PRO 4TB in RAID 5 along with a Samsung 970 Evo Plus 256GB NVMe drive as cache

  • The DS920+ sounds like a good option, I recommend you to take a look at the sizing guide document for Rapid Recovery, you can find it here Rapid Recovery - Technical Documentation (quest.com). It has additional details regarding the storage and things to consider. 

  • Just adding support to your thought process, yes this will do just fine. Try not to go below 16GB and stay clear of USB disks for repos. Basically any Synology or QNAP NAS will do nicely for your situation. A glorified desktop will be sufficient yes, like the Dell Optiplex line always made for decent DP hosts. Cheers. 

  • Thanks.  I actually got the Synology up and running yesterday (using the same server for now). Because some of our older servers have lots of files that are no longer necessary, and the Synology has a bit less usable space than the old storage, I removed them from the core and added them back as new protected machines.  (I cleaned off gigs and gigs of data so I wasn't wasting space backing up useless data.  The question now is do I have to do the same thing for our ERP server?  It looks like there is no way really to migrate just the backups for that one server since RR contains all the backups in one large file. I have the backups replicated to an offsite workstation, so I think that would give me some recovery points if I removed it and re-added it?   The thing is I need it to not point to the old storage.  I'm pouring over the knowledge base now to get some ideas

  • You can move the old data to the 'new' core/repo either by replication or by archive. The Replication requires 2 cores (old and new) and 2 repos (old and new) - once you replicate all the data over and its in-sync you can add the protected node to the 'new' core (after you remove replication and keep the recovery points) and it'll pick up and do incremental RPs. If you go with archive, you have to archive the data off the old repo/core and then import it on the new core/repo, however if you don't want a new base (full) backup you have to that while you have backups paused otherwise there will be a break in the RP chain between the old and the new, and thus get a new base/full backup.