Master Node election not working

[jijo.john]

Hi Team,

We are running a POC on elastic search. We have three nodes

Node 1:
node.master: true
node.data: true
node.ingest: true
node.ml: false
cluster.initial_master_nodes: ["ibs-dc-siem-1", "ibs-dc-siem-backup"]

Node 2:
node.data: true
node.master: true
node.voting_only: false
node.ingest: true
node.ml: false
node.attr.box_type: hot
cluster.initial_master_nodes: ["ibs-dc-siem-1", "ibs-dc-siem-backup"]

Node 3: (Coordinating Node)
node.data: false
node.master: false
node.ingest: false
node.ml: false

If Node 1 Goes down , Kibana on all three goes down and Node 2 will not become master

If Node 2 goes down , Kibana will be available from Node 1, Node 2 and Node 3 all works fine.

Below is the error we get from kibana

`{"statusCode":500,"error":"Internal Server Error","message":"all shards failed: [node_not_connected_exception] [ibs-dc-siem-1][10.181.10.176:9300] Node not connected"}

Output When all three nodes are UP

GET /_cat/nodes
10.181.10.176 5 99 19 1.61 1.15 0.76 dim * ibs-dc-siem-1
10.181.10.12 60 98 7 0.52 0.28 0.28 dim - ibs-dc-siem-backup
10.181.10.5 20 89 0 0.11 0.09 0.11 - - ibs-dc-siem-cn

GET _cat/master
6Kc_4PJcTyKUPeAzvKRUCQ 10.181.10.176 10.181.10.176 ibs-dc-siem-1

Is there any more config to be done . Please help to resolve this

[Christian_Dahlqvist]

I would recommend you read this. Just because you can make different nodes have different roles does not mean this is a good idea in a small cluster. I would recommend that you set up all nodes with default configuration and start from there. Your current configuration does not make much sense to me and I do not understand what you are trying to achieve or learn from it.

[jijo.john]

Hi Christian,

Thank you very much for pointing this document. I was able configure the nodes as per the doc and its worked perfectly. I am new to elasticsearch and we are doing a POC with small cluster with few clients pointed to it.

Thanks,
Jijo John