DR for VROPs

[jschaben]

Currently we have two DCs, each with a VC.  The first vROPs appliance was deployed to the DC that was closest  to us and then pointed to both the local DC's VC and the remote DC's VC. Allowing both DC's to be monitored from a single pain of glass. Now we have a requirement that if our current  DC where vROPs is, becomes a smoldering hole in the ground they want the other DC to not only monitor both DC's but also have the historical data available.  We tried standing up vROPs in the remote DC, but when we went to go register one of the VC's it told us that it was already being monitored by another vROPs instance. I am going to assume if we processed the original vROPs Appliance  will stop monitoring the VC. If this is not true please let me know.

Keeping in mind the following restrictions, does anyone have any ideas on how to make VROPs redundant, HA or DR minded accross two DC's?

- No Storage replication between sites.

- Each site has its own separate network.

Thank you in advance for any information, and apologies if these has already been mentioned in the forums.

[rcporto]

Check if helps: http://imallvirtual.com/creating-vrops-high-availability-cluster-part-1/

[MichaelRyom]

Hi

You get the warning because a vCenter only supports a 1:1 relationship with a vRops. Where as vRops support 1:many. The reason there's a 1:1 releastionship, is that when you in the webclient click action on an object and choose the open in vrops options, there can only be one vRops.

The last registered vRops to a vCenter gets the option for the vCenter. But it can be overridden manually.

But you can just contiue from the warning and both vRops will colled data. So that is not a problem.

There is the option to enabled HA, but placing a vRops node i each datacenter isnt supported afaik and there are some very string guide for supporting a vRops platform, in terms of latency. See the sizing guide for that.

Hope this helps.