alexus
December 11, 2017, 8:26pm
1
I have elasticsearch 6.x cluster and I just noticed that status is yellow
GET _cluster/health
{
"cluster_name": "X",
"status": "yellow",
"timed_out": false,
"number_of_nodes": 5,
"number_of_data_nodes": 5,
"active_primary_shards": 1911,
"active_shards": 3824,
"relocating_shards": 0,
"initializing_shards": 0,
"unassigned_shards": 1,
"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": 99.97385620915033
}
How does one troubleshoot? How can I make it green again?
Please advise.
dadoonet
(David Pilato)
December 11, 2017, 8:46pm
2
You probably need to wait until all shards are recovered.
BUT you probably have too many shards per node. You should reduce that number.
alexus
December 11, 2017, 8:50pm
3
I did noticed that one of nodes was down and I restart it and even though I'm not seeing any "Shard Activity", cluster health still says yellow(
dadoonet
(David Pilato)
December 11, 2017, 10:08pm
4
alexus
December 11, 2017, 10:19pm
5
@dadoonet
cat pending tasks | Elasticsearch Reference [6.0] | Elastic
# curl --silent --request GET --header 'Content-Type: application/json' $ELASTICSEARCH_URI/_cat/pending_tasks?v
insertOrder timeInQueue priority source
#
Cluster Allocation Explain API | Elasticsearch Reference [6.0] | Elastic
# curl --silent --request GET --header 'Content-Type: application/json' $ELASTICSEARCH_URI/_cluster/health/explain?pretty
{
"cluster_name" : "X",
"status" : "red",
"timed_out" : true,
"number_of_nodes" : 5,
"number_of_data_nodes" : 5,
"active_primary_shards" : 0,
"active_shards" : 0,
"relocating_shards" : 0,
"initializing_shards" : 0,
"unassigned_shards" : 0,
"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" : 100.0
}
#
this time it said "red", but I'm looking in Kibana and it says yellow
alexus
December 12, 2017, 8:57pm
6
possibly related to: Kibana Monitoring App reports Elasticsearch as Version: 6.0.0 instead of 6.0.1 - X-Pack - Discuss the Elastic Stack
as I restarted one of the nodes that was 6.0.0, now all nodes are on 6.0.1 and there are no unassigned_shards
anymore, and status back to GREEN!
system
(system)
Closed
January 9, 2018, 8:57pm
7
This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.