We recently tried to upgrade from 7.17.0 -> 7.17.8 in a rolling way, but in the middle of the upgrade we found replication error.
explanation" : "cannot allocate replica shard to a node with version [7.17.0] since this is older than the primary version [7.17.8]"
This wasn't expected for a patch upgrade like this as no significant change to indexing itself should be there. Can anyone explain why this is the case even for a minor/patch upgrade?
If I'm not wrong this is expected and documented.
The rolling upgrade section have this information:
Running multiple versions of Elasticsearch in the same cluster beyond the duration of an upgrade is not supported, as shards cannot be replicated from upgraded nodes to nodes running the older version.
Also, the first recommended step before the rolling upgrade process is to disable the allocation of replicas so replicas will not move arround during the rolling upgrade.
You should just finish the upgrade.
Thank you for your response. Yeah, I see this but actually wasnt expecting shard allocation to be a problem for such minor version change as well. This definitely makes sense for feature upgrades and major version changes. But I guess I have to stick with the recommendations to avoid replication errors for any upgrade.
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.