This blog post describes the steps to follow to scale up a 2-node HA enabled VMware vROps cluster:
Start Situation:
- SMALL1 and SMALL2 are deployed and are configured in HA mode
- SMALL1 is the Master node
- SMALL2 is the Master Replica node
Process to scale-up 2 HA enabled nodes
- Deploy 2 additional nodes (MEDIUM1 and MEDIUM2)
- Add MEDIUM1 and MEDIUM2 to the HA cluster as additional data nodes.
- Take the master replica node (SMALL2) offline
- Remove the master replica node (SMALL2) from the HA cluster
- Select one of the newly added nodes (MEDIUM1) as the new master replica
- Take the master node (SMALL1) offline
- Remove the master node (SMALL1) from the cluster
- vROps will promote node MEDIUM1 as its master node
- Select the second new node (MEDIUM2) as the new replica node.
Notes:
- During this process the cluster will restart several times, so dataloss is to be expected.
- This migration process is not officially documented by VMware, but is currently the only way to scale-up a VMware vROps environment. VMware advises a well-thought sizing before deploying any VMware vROps nodes.
- Having a cluster with mixed node sizes is not recommended and should only occur during a scale-up process
- In order to make an accurate sizing estimation of your environment, VMware designed a spreadsheet to calculate the vROps environment size. The link to download the spreadsheet can be found in VMware KB 2140551