Regressing vROPs Upgrade

[sandsfootgroup]

Hello

We are looking at documenting vROPs upgrade 6.0 to vROPs 6.6.1 (2 step 6.3.1 first).  Before any upgrades we will snapshot the vROPs server (only 1).  We have been asked to provide regression steps.

If we wanted to go back to, could we simply revert to the snapshot image?

Thanks

[mghall]

We've had to revert doing exactly that in the past. (We just recently had to back out a failed upgrade from 6.61 to 6.7 by using the snapshot and had no problems.) I see no reason why it should work from 6.0 to 6.61

[daphnissov]

The only precaution to take is to ensure you stop the vROps cluster before you take the snapshot. In that case, there are no consistency concerns with any of the databases.

[silus]

I realise this is a little late but just in case it might help someone else.

I found it best to shut down the vrops cluster from the admin console and also shut down the VM and then snapshot it.

The reason for this is even with the cluster shut down, if the VM is up, if you have to revert to snapshot when you boot it up the OS complains of filesystem errors and you then have to go in and run fsck on sda3 to get it back up. This can be avoided by simply snapshotting the shut down vm. Then you can power it on, the cluster will remain offline as you offlined it in the admin console, and then just continue the upgrade.