CCS requires wildcard after index name to search or always returns 404

Elastic Stack Elasticsearch
1

Quick bit about the setup:
Dedicated CCS cluster (no local indices, except for monitoring) v7.17.10
Two clusters clusterA and clusterB are connected as remote clusters v7.4.1

All of these searches are performed on the CCS cluster:
GET /clusterA:index-alice-2023.11.26/_search - Returns results as expected
GET /*:index-alice-2023.11.26/_search - Returns:

{"error":
  {"root_cause":[
  {"type":"index_not_found_exception",
   "reason":"no such index [index-alice-2023.11.26]",
   "index_uuid":"_na_",
    "resource.type":"index_or_alias",
    "resource.id":"index-alice-2023.11.26",
     "index":"index-alice-2023.11.26"}
  ],
"type":"index_not_found_exception",
"reason":"no such index [index-alice-2023.11.26]",
"index_uuid":"_na_",
"resource.type":"index_or_alias",
"resource.id":"index-alice-2023.11.26",
"index":"index-alice-2023.11.26"
},"status":404}

GET /*:index-alice-2023.11.26*/_search - Note the additional wildcard on index name, this returns the results as expected ...

I can't think what is causing this. Remote settings are

{
  "clusterB" : {
    "connected" : true,
    "mode" : "sniff",
    "seeds" : [
      "seed.clusterB.com:9300"
    ],
    "num_nodes_connected" : 3,
    "max_connections_per_cluster" : 3,
    "initial_connect_timeout" : "30s",
    "skip_unavailable" : false
  },
  "clusterA" : {
    "connected" : true,
    "mode" : "sniff",
    "seeds" : [
      "seed.clusterA.com:9300"
    ],
    "num_nodes_connected" : 3,
    "max_connections_per_cluster" : 3,
    "initial_connect_timeout" : "30s",
    "skip_unavailable" : false
  }
}
2

Seaching through queries from packetbeats, I see that almost every user is including ignore_unavailable=true in URL :smiley: This makes sense now. And the wildcard is allowing it to work without that set.

3

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