Disaster recovery for ESXi

[Zammo]

Hello guys,

 

I have been looking into a disaster recovery solution for our site. At the moment we backup our virtual servers in the "traditional" way.... to tape, file by file incremental with a weekly full.

 

I have been looking at using VEEAM to back up from our SAN located in the server room to a backup server located at the other end of the school. I believe something like this should be able to host the VEEAM software and store our backups: QNAP TS-879U-RP 24TB 8 Bay Rackmount iSCSI NAS (TS-879U-RP/24TB) - www.misco.co.uk.

 

The obvious weakness is that we are still reliant on our ESXi hosts, also located in the server room. How do people get around this? I was thinking of purchasing a small server, locating it elsewhere and installing ESXi with the idea it can be fired up and our VMs imported from the VEEAM backup in the event of losing both the SAN and the ESXi Hosts. Is this the best way of doing this? If so what sort of compatibility issues might I come across (processor?)

 

Thanks in advance

[Davit2005]

You shouldn't encounter any issues in theory, the ESXi versions may have to be the same.

 

After all the VM should be Virtually hardware independant

 

I use the Free version at home and many times have used it to move VM's to different hosts by the backup/restore method when I've wanted to completely re-jiggle things.

 

We use Veeam here for last 2 years with a SAN/Blade setup and are very happy with it and largely the support has been great.

Edited February 7, 2014 by Davit2005

[synaesthesia]

Theoretically you shouldn't have any, the guests don't care what they're on as long as it's ESXi and the host doesn't matter as long as drivers are all in place. But in reality it's no different to a physical server. You wouldn't have an identical physical box spare for every physical server so the same applies here. As long as other measures are in place like hotswap drives for the RAID arrays/SAN arrays, redundant power supplies etc you should be OK. Processors, mainboards, memory etc don't tell to fail so you don't need to keep those on hand so much. If they do it's usually down to external factors out of your hands. There's only so much you can prepare for. VEEAM is an awesome backup system.

[Zammo]

Thanks guys

 

So just to confirm, if I set up regular VEEAM backups to the QNAP, and then had something like this: Hewlett Packard ProLiant DL385p Gen8 Rack Server (F0B21A) - www.misco.co.uk in waiting, with ESXi installed, I could be back up and running with our core services just as quickly as I can copy the VMs across?

 

(please don't judge me too closely on the server I linked to, I just pulled one out of Misco's website)

 

Thanks!

[teejay]

Your plan makes sense, but you still need to have some means of taking a backup off-site, either by backing up to the cloud or to tape.

[teejay]

If you want really quick recovery, look at VMWare Site Recovery Manager, but that is complicated and a bit of a PITA.