Unassigned shards after node was restarted

Elastic Stack Elasticsearch
1

I am using v7.5.2, I have 10 nodes, 6 Data nodes having 8 TB of data and there is plenty of space left.

I have an ILM policy to rollover index after a month from Hot to warm.

I restarted a node after minor config change and I have 60-80 unassigned shards.
I have tried this API to allocate the replica.

POST _cluster/reroute

 "allocate_replica": {
           "index": "prod-2020.04.23-100077",
           "shard": 0,
           "node": "10thnode"
}

I get this error constantly, spent whole day to resolve without much help.

{"error":{"root_cause":[{"type":"remote_transport_exception" ...
...

[NO(node does not match index setting [index.routing.allocation.require] filters [data:"warm"])]

...

Please note I have displayed only the "NO".

Any help please?

2

An unassigned shard means the master can not find any shard copies after the restart!

3

The reason for Unassigned is because of "NODE_LEFT" not because master didn't find any copies.

I found out from the "_cat/shards" API.

4

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