michael81877 commented on PR #6560: URL: https://github.com/apache/nifi/pull/6560#issuecomment-1286981083
I tested this code change with a modified version of 1.18.0 and 1.19.0-SNAPSHOT with a basic 3-processor flow. 1. I deployed a 3-node modified 1.18.0 cluster and let it come up gracefully 1. I updated the nifi image version in my deployment to the modified 1.19.0-SNAPSHOT and upgraded each node 1 by 1. Each upgraded node joined the existing cluster gracefully, which definitely wasn't the case prior to this code change. 1. I changed the nifi image version in my deployment back to 1.18.0 and performed another rolling upgrade. Each node again rejoined the cluster gracefully. This is a huge feature that enables zero downtime rolling upgrades. There may be edge cases, for example with ghost processors, but the flow stabilizes when the upgrade is complete. There may be an argument to make this configurable via `nifi.properties`, but this LGTM overall +1 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
