Highlighted
Contributor
Contributor

ElasticSearch health red after upgrade to 19.03

Hi! I was wondering if anyone had any tips regarding this issue;

The ElasticSearch service health reads as "red" after the VIDM upgrade from version 3.2 to 19.03 was done.

It seems that the error is due to unassigned shards, but I'm having trouble getting the exact error as the index seems to be wrong.

Here is some output:

node-02:~ # vamicli version --appliance

Version - 19.03.0.0 Build 13322314

Description - To update the appliance​

node-02:~ # curl -XGET localhost:9200

{

  "name" : "Eliminator",

  "cluster_name" : "horizon",

  "version" : {

    "number" : "2.3.5",

    "build_hash" : "90f439ff60a3c0f497f91663701e64ccd01edbb4",

    "build_timestamp" : "2016-07-27T10:36:52Z",

    "build_snapshot" : false,

    "lucene_version" : "5.5.0"

  },

  "tagline" : "You Know, for Search"

}

node-02:~ # curl -XGET localhost:9200/_cluster/health?pretty

{

  "cluster_name" : "horizon",

  "status" : "red",

  "timed_out" : false,

  "number_of_nodes" : 6,

  "number_of_data_nodes" : 6,

  "active_primary_shards" : 2013,

  "active_shards" : 4026,

  "relocating_shards" : 0,

  "initializing_shards" : 0,

  "unassigned_shards" : 34,

  "delayed_unassigned_shards" : 0,

  "number_of_pending_tasks" : 0,

  "number_of_in_flight_fetch" : 0,

  "task_max_waiting_in_queue_millis" : 0,

  "active_shards_percent_as_number" : 99.16256157635468

}

node-02:~ # curl -XGET localhost:9200/horizon/allocation/explain?pretty

{

  "_index" : "horizon",

  "_type" : "allocation",

  "_id" : "explain",

  "found" : false

}

node-02:~ # curl -XGET localhost:9200/_cluster/allocation/explain?pretty

{

  "error" : {

    "root_cause" : [ {

      "type" : "index_not_found_exception",

      "reason" : "no such index",

      "resource.type" : "index_expression",

      "resource.id" : "_cluster",

      "index" : "_cluster"

    } ],

    "type" : "index_not_found_exception",

    "reason" : "no such index",

    "resource.type" : "index_expression",

    "resource.id" : "_cluster",

    "index" : "_cluster"

  },

  "status" : 404

}

Any advice? Thanks!

0 Kudos
1 Reply
Highlighted
Contributor
Contributor

You should manually reallocate the shards.

Check this article:

https://www.janhos.be/vmware/elasticsearch-status-in-red-after-full-restart-vidm-cluster/

0 Kudos