I'm planning to build ES Cluster with dedicated Master, Coordinate, Ingest and Data nodes.
(I didn't separated roles in my first ES Cluster, and it became quite slow over time.)
Member of cluster will be like this:
2 Master nodes
1 Coordinate node
1 Ingest node
3< Data nodes (Probably with Nginx Load Balancing)
When my plan reached here, I got several questions.
So I have dedicated Master/Coordinate/Ingest/Data nodes now. Then, which node do I have to tell to Kibana and Logstash's configuration to communicate with my ES Cluster?
If I throw bulk input into one of my master node, does it goes to Ingest node and stores to data nodes automatically?
I cannot understand role of coordinate node very well. I think it's quite similar to master node, but there must be kind of difference between them, but cannot figure out clearly. Could anybody explain about coordinate nodes?
1 - You will need to configure Kibana and Logstash to "discuss" with your Elastic CLIENT, Master node only "coordinate" any trafic from/to Data. And Data nodes are only used for Data.
2 - Yes he will, in caes you successfully configured your nodes to be in the same cluster. All your bulk need to pass through the CLIENT to be ingest correcly (if i'm not wrong)
3 - Coordinate = Ingest Node (Client) "There is indeed no change in functionality from what was previously referred to as a "client-only node", however starting in 5.0 you have to disable ingest for a node to be coordinating-only, and starting in 5.3 you have to disable a node from executing cross-cluster search requests for a node to be coordinating-only"
Note : You don't need Nginx as LB for your Data nodes, Elastic can do that for you if you configured it
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.