HOWTO going back on production after using replica

[olivier.g]

Hello

I have 2 ESX 4.0 servers (ESX1 and ESX2) with inside disks storage (no shared SAN or NAS).
ESX1 server is the production one, ESX2 is the local hardware backup. I do replication jobs on my VMs as often as possible (15 minutes).

When my ESX1 server will go down (let say both power supplies are dead and I don't have spare ones), I will start the replica VMs on my ESX2 host, losing 15 minutes of datas (not a big problem).

But After few hours/days (when my power supplies will be replaced), how can I go back to the ESX1 host easily ? I don't want to lose days of new datas on the replicas

The first solution I see is to make a full backup of all VMs on the ESX2 and do a restore on ESX1 (erasing old ones on ESX1), but it significates a quite long outage for users because I will have to stop the VMs, backup them and restore them on ESX1.

How do you manage this ?
I'm sure there is a better solution. I don't have vMotion and all other cool VMWare stuffs to be able to move VMs from one ESX to an other instantly.

Thanks a lot for your help
Olivier

[Vitaliy S.]

Hello Olivier,

Please have a look at the thread below discussing the same question:

http://www.veeam.com/forums/viewtopic.p ... back#p5974

Thank you!

[olivier.g]

Sorry for not seeing this post, I have searched but..

I also have imagined this solution but think there was something better than myVM_replica_replica_replica_...

Thanks
Olivier

[matthadley]

Oliver - Good morning! I have the same setup as you and have made the same inquiries. It appears that a "quick" switchback/failback is something that Veeam plans to add in a future release.

There is some very good information about how to "commit" the data collected during a Veeam-managed failover in the following thread, titled "Data LOSS - Undo Failover": http://www.veeam.com/forums/viewtopic.p ... Y+SNAPSHOT

On a separate note: Are you running vSphere? Using vCenter? If so, do you have it in a VM? Do you run Veeam Backup and Replication in the same VM? Just curious how others with multiple physical servers/direct attached storage are running their VM and Veeam environment.

I am in the process of deploying VMware vSphere and Veeam Essentials using two Dell R510s with (8) 1TB Sata drives and 32Gb of RAM. I plan to run one MSSQL production server, a few other related servers, and a console (vCenter/Veeam) guest. I plan to put the console guest on what you refer to as ESX2 and replicate guests from ESX1 to ESX2. I figure that if ESX1 is my primary and it fails, the vCenter/Veeam will be running on my standby and all is good. If ESX2 fails, my primaries are still running -- still good. *RESARCHING THE FOLLOWING* (I plan to replicate my vCenter/Veeam console in reverse (ESX2 to ESX1) so that I'll have a safe copy of it, too, good to about 15 minutes of age. ) -- there are threads on ways to do this but I haven't gotten through them yet.

All that said, after a core failure of ESX1, recreating an ESX1-to-ESX2 model, of the guests will be tedious until Veeam adds the aforementioned quick switchback functionality.