I am bit confused about first 2 steps. Its said that stop indexing if possible. What is relation between first step and second step.
Where should I set the transient variable before I shut down node 3? on master node I guess. right?
When I shut down and restart node-3, during that time, will that data be indexed only on node-1 and node-3? I hope indexing will not have any blockage during the maintenance/unavailability of node-3. Am I correct?
Whichever node. It will be routed to master anyway.
You meant node-2 right?
So yes your data will be still indexed on the other nodes.
But some replicas won't be get the index operations as they were allocated on the node you just shutdown.
That's one of the reasons we recommend stopping indexing while doing that. Recovery process will be much faster if all shards are still synchronized.
But, if I want to reboot Node-2, what is the use of disabling routing allocation on that node?
Is it like...
I will disable routing allocation
Then I will restart elasticsearch.
Then it will route to master only.
Then I can reboot node-2
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.