vApp Failover Boot Order

[jbarrow.viracoribt]

I have been asked to put together a strategy for failing over vApps from our primary datacenter. Many of our vApps are configured for specific boot rules. For instance, one vApp we have boots the SQL server first, makes sure VMWare Tools is good, then it allows the other two "app" servers within the same vApp to boot. If I ran a failover using Veeam but picked this vApp, would I be able to choreograph the same boot process or will it read it from the vApp? Right now at the DR site replica VM's are just in one long list, not grouped into vApp's like our PROD site.

Thanks in advance.

[veremin]

Just to doublecheck - we’re talking here about vCenter, not a vCloud one, right?

If I ran a failover using Veeam but picked this vApp, would I be able to choreograph the same boot process or will it read it from the vApp?

You will have to do it either manually or write a custom script, since VB&R isn’t vApp-aware.

Thanks.

[jbarrow.viracoribt]

Yes, vCenter not vCloud.

Thanks for the response.

[jbarrow.viracoribt]

How about this. What if I recreated my vApp's on the DR site where my replica VM's exist, adding the equivalent replica VM's to vApp's that match PRO. Then, if I used Veeam to failover that vApp, could I tell Veeam to not power on the VM's but to let me do that manually? Then I could log into the vCenter console at my DR site and power on the vApp manually, that would honor the vApp rules. Would that process work or does that mess up what Veeam would be trying to do for me?

[veremin]

Then, if I used Veeam to failover that vApp, could I tell Veeam to not power on the VM's but to let me do that manually?

Actually, not. The whole idea of failover is that replicas should be powered on.

If you power on these VMs either manually (via vSphere client) or using vApp specific rules, they would work fine. Nertheless, replication jobs would be affected, since VB&R would stay anaware about changes that have happened. For instance, VB&R wouldn't be able to perform failback, etc.

Thus, it’s always recommended to have VB&R do the required operations for you. The available options are:

Manual way.

1. Manually failover SQL VM.
2. Check VMware tools status.
3. Manually failover remaining VMs.

Automatic one. In this case PS script would do virtually the same.

Thanks.