SAN volume replication vs Veeam backup replication

[cparker4486]

Hello!

I'm considering doing some remote backup to a datacenter and the owner there is trying to get me to replicate my SAN volume(s) to his SAN instead of replicating Veeam backups. I don't having pricing yet (so that factor is not in the picture) but I wonder if there are any general pros and cons to either method?

One con I think of to replicating SAN volumes is that I may have to pay for more storage space than I really need because my volumes don't contain only my Tier 1 VMs. That is, I will have to replicate more than one volume to protect my Tier 1 VMs. Conversely if I went with replicating Veeam backups I could package up all my Tier 1 VMs together.

I'm a bit in the dark on this so any comments are appreciated.

[Helqasem]

Send them this white paper
http://go.veeam.com/disaster-recovery-2 ... em-en.html

Alternatively the recorded webinar:
http://www.veeam.com/videos/five-most-c ... -1441.html

Hope this helps.

[cparker4486]

Yes! This is exactly what I was hoping to find! Just from reading the bulleted list there I can see that "replication of corrupt data" is a perfect example.

[Gostev]

I always like to be upfront and honest about SAN replication, because it is truly awesome DR strategy that beats our offering in certain scenarios.

SAN replication really is great if you can afford it (both hardware and required bandwidth will be very expensive). It does provide big bang for the buck though - specifically, lowest RPO for certain specific types of disasters. But, because you want to be protected from ALL types of disasters, you still need to perform normal backups or replicas (depending on RTO requirements) to the 3rd storage regardless - even if you are doing SAN replication.

Another way to look at this is:
- Standard offsite backups or replicas to independent storage are MANDATORY, you should do them no matter what.
- SAN replication is optional EXTRA that beats standard backups or replicas in terms of RPO in specific datacenter-wide disasters. Low RTO is also achievable with SAN replication, but it will require additional expensive 3rd party product (such as VMware SRM) to orchestrate failover/re-IP (something that is included with Veeam replication).

There is one horrible disaster story that I happen to know first-hand. I will just say, it is a very big name in Las-Vegas. They had their entire disaster recovery strategy based around SAN-based replication. They were hoping for DR site to have a good copy of their production in case of disaster. And it actually worked exactly as it should: what they have found there when storage disaster struck were perfect copies of production LUNs, including all the corruption. Obviously, bad data replicates just as well as good data.

Ironically, we tried to sell them product last year, but they were absolutely convinced (by their storage vendor) that their SAN snapshot based data protection strategy was flawless, and did not need proper backups on top of that. They really bought into that "no backup windows" speech, guess it sounded way too appealing after their previous experience with VM backup solution (one well-known Veeam competitor that is designed not to fit backup windows).

Lots of IT people were laid off there (sadly, including our champion, who actually pushed everyone to implement Veeam). People were laid off because company lost hundreds of millions USD during a very extended downtime. There are no possible excuses for these kind of losses...