I was looking through some documentation and it looks like as long as you have a failover node you can remove the primary node from the clsuter and replace it without losing data is that correct? Are there any other considertaitons to make. The Prmiary in my cluster is running windows 2008 r2 and I want to switch to a newer OS. So my plan is to have a new server built. Failover to my other node. remove my currenlt primary node. Replace with the new server.
Anything else I need to include?
Can you clarify your intentions for "not losing data"? The primary and failover nodes do not replicate data, so there would be no historical data on the new host other than what could be reprocessed from the AMDs when the new host is brought online. The retained failover node would still have the historical data, but that data would generally be unavailable unless you were specifically running reports from the failover node, in a failed over state, or promoted the failover node to primary and added the new host as a replacement failover node.
I know that when you mess with the devices in your cluster you can lose some historical data, but since the Failover node should have the same data as the Primary I wondered if I failed over to the failover node, and replaced my primary with a new server with that cause an issue with the failover nodes data also, or just the database for the Primary.
I am really just trying to understand what data impacts come from making changes in the system, and is there a way to avoid losing historical data.
To retain your data, the best method would be to move the existing database from the old server to the new server as part of the replacement of the server. This will retain all of the data, while still giving you a path forward to replace the server' chassis and/or OS.