VMware Cloud Community
OsburnM
Hot Shot
Hot Shot
Jump to solution

How to rebalance resource count after adding more data nodes

Preparing to upgrade vrops to 8.2 and the upgrade tool gave me a red splat that I have nodes with high resource counts.  After re-running the sizing calculator, we realized that adding all our DellEMC storage instances far exceeded our counts.  So we right-sized the cluster (added more data nodes) per the sizing calc sheet and ran the adapter rebalance; however, the UAT pak is still saying we're in excess.  It doesnt seem that adding additional nodes changed the original resource count values.  

Is there something else we are supposed to do to get the resources moved to new data nodes so it's more balanced?  I've got three data nodes with over 29k objects each yet the new ones only have 500 each.  What am I missing?

***UAT Message***

Validating the number of resources per node meets the sizing requirements.

Node *nodenameFQDN* has resource count of 29394.
Validate the product sizing scaling spreadsheet to ensure the product is sized with the correct number of nodes to support the scale of resources. (https://kb.vmware.com/kb/2093783)
**********************
 
Reply
0 Kudos
1 Solution

Accepted Solutions
KabirAli82
Expert
Expert
Jump to solution

From the docs;

https://docs.vmware.com/en/vRealize-Operations-Manager/8.1/com.vmware.vcom.core.doc/GUID-3214342C-AC...

Adding Data Nodes

Data nodes are the additional cluster nodes that allow you to scale out vRealize Operations Manager to monitor larger environments.

You can dynamically scale out vRealize Operations Manager by adding data nodes without stopping the vRealize Operations Manager cluster. When you scale out the cluster by 25% or more, you should restart the cluster to allow vRealize Operations Manager to update its storage size, and you might notice a decrease in performance until you restart. A maintenance interval provides a good opportunity to restart the vRealize Operations Manager cluster.

In addition, the product administration options include an option to re-balance the cluster, which can be done without restarting. Rebalancing adjusts the vRealize Operations Manager workload across the cluster nodes.
 
 
If you're 100% sure that your new setup can support the load, I would go ahead and open a support case as to why the rebalancing isn't working (properly).

Was I helpful? Give a kudo for appreciation!
Braindumping @ http://kablog.nl/
Tweeting @ https://twitter.com/_Kabir_Ali_

View solution in original post

Reply
0 Kudos
3 Replies
KabirAli82
Expert
Expert
Jump to solution

From the docs;

https://docs.vmware.com/en/vRealize-Operations-Manager/8.1/com.vmware.vcom.core.doc/GUID-3214342C-AC...

Adding Data Nodes

Data nodes are the additional cluster nodes that allow you to scale out vRealize Operations Manager to monitor larger environments.

You can dynamically scale out vRealize Operations Manager by adding data nodes without stopping the vRealize Operations Manager cluster. When you scale out the cluster by 25% or more, you should restart the cluster to allow vRealize Operations Manager to update its storage size, and you might notice a decrease in performance until you restart. A maintenance interval provides a good opportunity to restart the vRealize Operations Manager cluster.

In addition, the product administration options include an option to re-balance the cluster, which can be done without restarting. Rebalancing adjusts the vRealize Operations Manager workload across the cluster nodes.
 
 
If you're 100% sure that your new setup can support the load, I would go ahead and open a support case as to why the rebalancing isn't working (properly).

Was I helpful? Give a kudo for appreciation!
Braindumping @ http://kablog.nl/
Tweeting @ https://twitter.com/_Kabir_Ali_
Reply
0 Kudos
OsburnM
Hot Shot
Hot Shot
Jump to solution

Okay, I absolutely feel foolish.  For one, I skipped right over the part about taking the cluster offline/online when increasing by 25% or more... but that's not the worst part.  I kept rebalancing the collector groups and not rebalancing the cluster.  😩  I'm calling a do-over for the week.

KabirAli82, thanks for the reply.  The tunnel vision got the better of me this time.

Reply
0 Kudos
KabirAli82
Expert
Expert
Jump to solution

No worries, happens to the best of us. Glad I could help.


Was I helpful? Give a kudo for appreciation!
Braindumping @ http://kablog.nl/
Tweeting @ https://twitter.com/_Kabir_Ali_
Reply
0 Kudos