vRealize Operations 6.2 size migration - Medium to Large

[AnkitCP]

Hi,

I have a six node vROPs cluster in which each node is of size Medium i.e 8 vCPU and 32 GB Memory.

Is it possible if I can convert the vROPs virtual appliances to size Large i.e.16 vCPU and 48 GB Memory, and then shrink the cluster to 4 nodes?

If yes, please suggest the procedure to achieve the same.

Thanks,

Ankit

[rcporto]

I cannot found any official documentation about scale up vROPs nodes, but take a look here: http://www.definit.co.uk/2016/04/vrops-what-if-i-want-to-scale-up/

Another option is to add new large nodes and than remove the old medium nodes, see: VMware vRealize Operations Manager – Scaling up a 2-node HA enabled vROps cluster » ITuDA

[sxnxr · Accepted solution]

I have up scaled 8 nodes from small to large simply by taking a node off line, shut it down and add the required CPU and memory.

vXpress: Scaling up vRealize Operations Manager Nodes!

I have never removed a data node before and only done remote collectors but take a look at

Removing a node from a VMware vRealize Operations Manager 6.x cluster (2107711) | VMware KB

[mark_j]

To increase the size of the nodes, you simply power them down, RC, Data, Master Rep, then Master.. add the resources (cpu/mem, etc), then bring them back up in reverse order. In terms of REMOVING nodes from the cluster, you are going to lose data - period. You can do it gracefully, bringing them offline, then removing them, however there is no way to gracefully "evacuate" perf data from the nodes. It's a known caveat to vR Ops, but if you want to remove the nodes you'll need to bite the bullet and lose some data points.

[AnkitCP]

I worked with VMware support to upgrade the vROPs nodes config from Medium to Large.

Also, the support person directed me to the same VMware KB article Removing a node from a VMware vRealize Operations Manager 6.x cluster (2107711) | VMware KB

He asked me to take snapshot of all vROPs nodes and said that there will be no data loss, as we also upgraded the disk size of vROPs nodes to accommodate for data .

But, eventually we dropped the plan to remove the vROPs nodes from cluster and kept it in view of future data growth.