4 ES nodes on single Server, who does what and whom to connect?


(hasaN khaN) #1

I have created 4 nodes of ElasticSearch on single server by using these properties

node_0 (Master Node)

node.master: true
node.data: false
node.ingest: false

node_1 (Data Node)

node.master: false
node.data: true
node.ingest: false

node_2 (Ingest Node)

node.master: false
node.data: false
node.ingest: true

node_3 (Coordinating Node)

node.master: false
node.data: false
node.ingest: false

My queries are

  1. To which node I connect from
    a) Kibana
    b) Logstash / Beats
    c) From Browser or ElasticHQ (admin/monitoring tool)

  2. and how can I combine master node and coordinating node

  3. If Kibana can connect to only 1 ES node then what if 1 node gets down and other nodes are up running fine.


(David Pilato) #2

I believe you are only doing that for tests, right?

Your settings are correct.

  1. a. node_3
    b. node_3
    c. node_3

  2. node_0 can actually play that role. So you don't need node_3 in that case.

  3. Kibana won't display anything anymore.


(hasaN khaN) #3

Thanks for your reply, I am planning this architecture for production setup and wanted to avoid downtime and split-brain/data loss situations.

What I understood from reply no. 2 is if node_0 is created as a master node then node_3 is not requiredand the answer for point no. 1 a,b,c will change to node_0? correct?


(David Pilato) #4

But you are planning this architecture on multiple physical machines right?

If you have only 4 nodes, then keep data and master on all nodes. You can if needed use the 4th node to be ingest (if you are using ingest pipelines - ingest name might be confusing I know) and no data, no master. A coordinating node that is.


(system) #5

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.