Comprehensive data protection for all workloads
Post Reply
Anonymouse
Lurker
Posts: 1
Liked: never
Joined: Dec 16, 2010 5:27 pm
Full Name: David
Contact:

SSD and vPower

Post by Anonymouse »

Is SSD ok to use for the c drive of a dedicated veeam server? P.36 of the user guide says that 100gb of free disk space is the minimum space for storing the VM recovery write cache.... so a 128gb ssd drive with just server 2008 and Veeam would technically qualify. (And a 256gb ssd drive definitely would..).

I'm just not sure if ssd is a good idea for Veeam server? My concern would be performance if I have to run a machine in instant recovery/vPower mode.. not so much the restore time... (and the backups are stored on sata drives, just OS/Veeam on the ssd drive...)


?

Thank you!
Gostev
Chief Product Officer
Posts: 31460
Liked: 6648 times
Joined: Jan 01, 2006 1:01 am
Location: Baar, Switzerland
Contact:

Re: SSD and vPower

Post by Gostev »

Performance will be acceptable with or without SSD. I am not sure if SSD will be able to improve performance significantly, at least based on my own performance testing with regular disks. In fact, SSD will likely be bad for performance in the long run, because of the infamous SSD slow down issue on non-TRIM aware operating systems (such as VMware hypervisors). So unless you have an SSD drive that provides alternative firmware with internal garbage collector (at least I know my SSD had that at some point), then I personally would not go for it. But if you do have one and decide to go ahead with this, I will be definitely interested to hear your results.

As for 100GB, don't think about it as a hard number. There are scenarios when 100GB will not be enough (running multiple heavy I/O VMs from extended period of times). For most typical usage scenarios however (instant recovery of one failed server followed by moving it back to production datastore shortly after) 10GB should be sufficient for most workloads. This is why 10GB is what we have always had listed in the System Requirements section of the Release Notes document as the recommended minimum. Of course, allocating much more disk space (for example, 1 TB) cannot hurt, and does give you more leg room in case of real disaster - you will be able to run instantly recovered VMs for weeks if you need to. This may be needed, for example, if you production SAN is dead and you simply do not have enough production storage capacity available to move instantly recovered VMs to. I saw this happening to other customers before... just no place to restore VMs from failed SAN to, literally.

Thanks!
Post Reply

Who is online

Users browsing this forum: No registered users and 208 guests