Question on 3-node resiliency with RF-2

Question

Hi. New to Nutanix. We have a 3-node test cluster to evaluate. We are at RF-2 (1N/1D). I am hoping someone can clarify some things for me.

I know I can lose 1-node or 1-drive and I will still have data availability. My question is this:

If I lose 1-node will data-resilency be rebuilt in that I could actually lose a drive in 1 of the two remaining healthy nodes and still have data available? My thought process is Curator will simply rebuild the data to another healthy drive on the same node?

I am trying this now where I cleanly shut a node down, and after 44 minutes Element still shows data-resilency as critical. I would assume since there is tons of free space the resilience would move to “OK” and I could tolerate a drive failure. I must really be misunderstanding something, I am certain I have read that would be work. I just cannot lose a drive with the node down during the rebuild process.

Thank you for clarification!

DrewPlaster · Answer

Jeff-Atl,For clarity, you have Redundancy Factor for the cluster and Replication Factor for the storage; both are in play along with the number of nodes in the cluster for the number of nodes and drives that could potentially be lost and still have a functional cluster.The good news is without the need to do "critical" thinking pun intended; you can easily review the number of each tolerated for your given cluster:If your cluster is not in a healthy state, click on the Critical status to get a list which component(s) do not support additional failures at the current state and which components can support failures.If your cluster is in a health state, click on the Redundancy State to review the selected Desired Redundancy Factor which includes # Node / # Drive failures tolerable.

Sign up

Login to the community

Scanning file for viruses.

This file cannot be downloaded