My cluster have 4 data nodes, and this caused many bulk rejections,we didn't have any update operate,Elasticsearch version is 6.4.0
GET _cat/shards?v&index=2019.01.22&h=index,shared,prirep,stat,node
index prirep node
xxx_accounting-2019.01.22 p jdB9njq
xxx_accounting-2019.01.22 r an8XFEI
filebeat-syslog-2019.01.22 r M8bpmYn
filebeat-syslog-2019.01.22 p jdB9njq
filebeat-syslog-2019.01.22 p jdB9njq
filebeat-syslog-2019.01.22 r an8XFEI
filebeat-syslog-2019.01.22 r CrBTwW6
filebeat-syslog-2019.01.22 p jdB9njq
xxx_messages-2019.01.22 r CrBTwW6
xxx_messages-2019.01.22 p jdB9njq
metricbeat-login-2019.01.22 p jdB9njq
metricbeat-login-2019.01.22 r an8XFEI
metricbeat-xcompute-2019.01.22 p jdB9njq
metricbeat-xcompute-2019.01.22 r an8XFEI
metricbeat-compute-2019.01.22 r M8bpmYn
metricbeat-compute-2019.01.22 p jdB9njq
metricbeat-compute-2019.01.22 p jdB9njq
metricbeat-compute-2019.01.22 r an8XFEI
metricbeat-compute-2019.01.22 r CrBTwW6
metricbeat-compute-2019.01.22 p jdB9njq
metricbeat-compute-2019.01.22 p jdB9njq
metricbeat-compute-2019.01.22 r an8XFEI
metricbeat-compute-2019.01.22 p jdB9njq
metricbeat-compute-2019.01.22 r an8XFEI
auditbeat-6.5.4-2019.01.22 r M8bpmYn
auditbeat-6.5.4-2019.01.22 p jdB9njq
auditbeat-6.5.4-2019.01.22 p jdB9njq
auditbeat-6.5.4-2019.01.22 r an8XFEI
the thread pool info
curl localhost:9200/_cat/thread_pool?v | grep jdB9njq
dB9njq analyze 0 0 0
jdB9njq fetch_shard_started 0 0 0
jdB9njq fetch_shard_store 0 0 0
jdB9njq flush 0 0 0
jdB9njq force_merge 0 0 0
jdB9njq generic 0 0 0
jdB9njq get 0 0 0
jdB9njq index 0 0 0
jdB9njq listener 0 0 0
jdB9njq management 1 0 0
jdB9njq refresh 1 0 0
jdB9njq search 0 0 0
jdB9njq snapshot 0 0 0
jdB9njq warmer 0 0 0
jdB9njq write 48 165 7591394