Status: yellow / unassigned_shards


#1

I'm trying to upgrade my elasticsearch cluster from 6.1.1 to 6.1.2 and even though there are no pending_tasks anymore (after I switched that node to 6.1.2):

# curl --silent --request GET $ELASTICSEARCH_URI/_cluster/pending_tasks?pretty
{
  "tasks" : [ ]
}
# 

it looks like there are some unassigned_shards and that's probably my cluster's status is yellow:

# curl --silent --request GET $ELASTICSEARCH_URI/_cluster/health?pretty | egrep '\"(status|unassigned_shards)\"'
  "status" : "yellow",
  "unassigned_shards" : 4,
# 

How does one deal with this scenario?

Please advise.


#2

I've tried: Cluster Allocation Explain API | Elasticsearch Reference [6.1] | Elastic and got following output:

# curl --silent --request GET $ELASTICSEARCH_URI/_cluster/allocation/explain?pretty | grep explanation | tail -1
          "explanation" : "target node version [6.1.1] is older than the source node version [6.1.2] and may not support codecs or postings formats for a newer Lucene version"
#

(Loren Siebert) #3

As @Clinton_Gormley once said, "you've got mixed versions of ES by the looks of it. You need to upgrade all of your nodes to the same version"


#4

I figure it out based on explanation from previous reply) Thanks anyways!


(system) #5

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