I found that if the network failure happens, the es nodes who lost the
master become master itself and then never come back.
I want the lost nodes can come back, otherwise i will lost some index
data and have to restart the lost nodes.
?
I found that if the network failure happens, the es nodes who lost the
master become master itself and then never come back.
I want the lost nodes can come back, otherwise i will lost some index
data and have to restart the lost nodes.
?
Hi,
This is quite broad topic (and I am not expert in it). May be you
should consider using shared gateway or increase number of index shard
replicas. By the way this topic has been discussed on ML many times,
usually they refer to network partitioning or brain split problems.
For example check the following thread:
http://elasticsearch-users.115913.n3.nabble.com/CAP-theorem-td891925.html
Regards,
Lukas
Dne sobota, 18. června 2011, Weiwei Wang ww.wang.cs@gmail.com napsal(a):
I found that if the network failure happens, the es nodes who lost the
master become master itself and then never come back.I want the lost nodes can come back, otherwise i will lost some index
data and have to restart the lost nodes.?
© 2020. All Rights Reserved - Elasticsearch
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant logo are trademarks of the Apache Software Foundation in the United States and/or other countries.