Shard in es cluster

Elastic Stack Elasticsearch
1

I had a single node default "Elasticsearch" cluster so far. It had one index and two alias associated with that
curl -s -X PUT $elasticsearchURL/$indexName?pretty -H 'Content-Type: application/json' -d '{"settings":{"index":{"number_of_shards" : 1,"number_of_replicas" : 1}}}'

As you see I have shards as 1. So far I had only single node. So I didnt bother about shards.
Now I have created three nodes cluster with two new nodes together with previous node. I guess same index will get replicated in other two nodes with single shard once it joined cluster

  1. should I increase shards when I move to cluster?
  2. What is the impact if I leave above as it is?
2

This is a; It Depends.

You can add more replicas, which is useful for both querying and redundancy. You can also look to use the _split endpoint to create more primary shards for the existing index, this is useful for indexing throughput.

Ultimately it comes to what sort of data you have, and what you do with it.

3

We use to parse over CSV data using logstash and store it in index using alias

elasticsearch
{
hosts => ["localhost:9200"]
index => "dashboard_write"
ilm_enabled => false
}

dashboard_write is the alias for our index dashboard-*. We use to create rollover index everyday

Will I face any performance impact in my cluster due to single shard?

4

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