@barvepan, First You need to install the elasticserach (same version ) on all the new nodes. Then you can add the new node in your existing cluster. You can refer the below link:
You can take snapshots and restore them its only way of backup your data in elasticserach. You need to check the compatibility also of your old elaticserach version and new elasticserach version.
As, I explained earlier, we have (elasticsearch 1.5.2) with default configuration i.e. one cluster with one node. We are planning to migrate it to one cluster with 3 nodes.
We decided to add new nodes 1 master and 1 data node. I assume its just configurations in elasticsearch.yml file. Just have few questions,
what value we should give to discovery.zen.minimum_master_nodes . We want 3 nodes cluster ( 2 master +data nodes and 1 only data node )
While doing this configuration and settings do we need to take backup of existing data. Is there any chance of losing data in all these configurations.
I assume, after adding new nodes to existing cluster, elastic search will take care of distributing data across all nodes.
As, I explained earlier, we have (elasticsearch 1.5.2) with default configuration i.e. one cluster with one node. We are planning to migrate it to one cluster with 3 nodes.
We decided to add new nodes 1 master and 1 data node. I assume its just configurations in elasticsearch.yml file. Just have few questions,
what value we should give to discovery.zen.minimum_master_nodes . We want 3 nodes cluster ( 2 master +data nodes and 1 only data node )
While doing this configuration and settings do we need to take backup of existing data. Is there any chance of losing data in all these configurations.
I assume, after adding new nodes to existing cluster, elastic search will take care of distributing data across all nodes.
Its value calculate using the formula N/2+1. Where N is the number of master node. To avoid from split brain problem we should have minimum 3 master node. Then discovery.zen.minimum_master_nodes value will be 2 (3/2+1=2).
Yes, you need to take the backup of your data. Because when you upgrade your ELK all data will be lost. I am not sure whether you can restore the elasticserach1.5.2 snapshots in latest elasticsearch or not. The latest elasticsearch version is 6.5.0.
Yes, when you have multiple data node in your cluster the elasticsearch itself take care of data distributing across all data nodes.
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.