I have an application in PHP.
It uses Elastic search for searching( search , insert) operations.
I want to speed my application. So ,Recently i have made the ElasticSearch to clustering concept (Totally 3 nodes- data nodes also : 03)
Initial Node : node-1
New nodes : node-2, node-3
In my application , i have configured php to conenct to one of the ES ( node-1) in my code.
I need to know,
1) Is it enough to point one of the ES clustering node among ( node1, node2 , node3 ) in the PHP configurations ?
2) Will itself do some internal process to split the requests based on load
3) What are all the internal process ( php -> ES clustering )?
4) Do i need to give the all the IP in PHP configurations
Or
5) I need to mention for insert it should go to node-1 for search it should go to node-2 like that for proper functioning / speed up the process
6) How do i know which request goes to which node ?
To answer questions for 1-4, you need to purely depend on your elasticsearch configurations. What things you pointed out will support in elasticsearch configuration. Elasticsearch supports master-data failover. Question 5 is quit complex for which you can simply point out master IP for balancing. Operation like insert, search, etc. will be done by ES itself based on the node availability. We need not worry about it. Could you brief me on Question 6?