Replication to DR site

[davidintouch]

I am looking for some detailed documentation for configuring replicaton to a disaster recovery site. I have been to the Veeam Backup & Replication Help Center, but looking for something a little more detailed. Could someone please point me in the right direction for this?

[Vitaliy S.]

Hi David,

You should find all the details in our Evaluator's and User Guides which are available under the Resources tab on the product web page. Did you have a chance to take a look at these guides?

BTW - could you please tell me how does your DR site look like? Do you plan to use replication jobs only?

Thanks!

[davidintouch]

Thanks for the update. One thing I need clarification on is regarding automatically having VMs failover at the DR site. I read the information here that you have posted in other topics: http://blogs.vmware.com/vipowershell/20 ... larms.html

From that scenario or setup, vcenter would have to reside at the DR location, correct? Otherwise if vcenter was at the production site, and a disaster occurred, no alarms would be generated thus triggering any scripts/actions?

Thanks

[tsightler]

Using that scenario vCenter would have to exist at the remote site, or vCenter's HA would need to be provided by a different mechanism. Many customers simply have a completely different vCenter at the DR site. Veeam itself does not have any built in ability to perform automatic VM failover so anything customers choose to implement in this regard is based on their own solution.

The challenge here is preventing "split-brain" where one error triggers an automated failover when in reality the orginal site is not in a failure state. Because of this automatic failover can be very dangerous if not implemented correctly and with much care. I've heard horror stories of automated failover starting VMs and remote sites and actually causing outages and other more complicated issues. In one case, a company ended up with users logged in to two different systems based on their region, users in one region were still using the original servers, while users in another region were connecting to the DR servers. This actually went unnoticed for almost an hour and was a major issue to clean up since users were performing transactions in both systems.

So basically, this is just a warning, implement automatic failover with much caution!

[J1mbo]

So basically, this is just a warning, implement automatic failover with much caution!

I concur absolutely with that. I also don't much like vSphere HA.