Elasticsearch maintenance with cluster.routing.allocation.enable=none

When doing a rolling upgrade, it says to set "cluster.routing.allocation.enable=none" and then flush.

  1. Is the documentation for 6.2 wrong in that it disables using persistent and then does transient? Should it not be both transient? I see that someone else had the same question and also in the 5.6 documentation it says to use transient for disabling and enabling allocation (Rolling Upgrade Documentation Incorrect?).

  2. How is this considered a rolling upgrade since we are essentially down when issuing the "cluster.routing.allocation.enable=none" command? I am not able to create any new indexes at this time and I'm only able to add documents to existing indexes. Do these index creates just fail and disappear with the error "no allocations are allowed due to cluster setting [cluster.routing.allocation.enable=none]"? From my testing, these documents that I try and create while my cluster allocation.enable is set to none is lost. Please tell me if I'm wrong.


In the documentation it says, "While you can continue indexing during the upgrade, shard recovery is much faster if you temporarily stop non-essential indexing and perform a synced-flush." but I'm not able to create any new indexes. Am I missing something?

could someone please respond to this?

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.