We have an existing elasticsearch stack running basic license, is it possible to add more nodes to the cluster which do not hold any data and only pass data from a location to existing data / master nodes with basic license?
I read somewhere that if nodes are added to same cluster then remote cluster is not applicable, to clarify my proposed nodes are in different datacenter however will be sending data to existing cluster.
Okay, could you advise what node type i should be using for nodes in datacenter B (currently whole els setup is in Datacenter A). I need these new nodes to be just adding data to existing setup and not store anything locally in datacenter B
In general I would not recommend splitting a cluster across data centers generally that is not considered best practice and can result in an unstable cluster.
It is possible to do this but you need guaranteed low latency... And solid connectivity.
Personally I do not recommend but it is possible and some users do it.
A little confused by this but sound like you are describing a coordinating role only... Which only accepts queries and index requests but stores no data.
But why not just send the request directly from data center B to the cluster in Data center A... In the first place?
I am trying to avoid multiple servers trying to connect to existing elastic in DC A and allow 1 more elastic node ( which should only consolidate and pass on request to existing stack ).
This way its easier to allow communication on 1 node rather than multiple nodes trying to connect across different DC
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.