node1 ingestor
node2 master, data
node3 master, data
node4 master, data
I have an 379.4gb Index, with 5 shards and 2 replicas, each shard has 25gb, so i have some points to clarify:
Will elastic get the whole shard while reading it or will just get the blocks who contains the docs ?
Should i put one mode data node and put my index to 6 primary shards ?
Your data will reside across the 5 shards across the three data nodes.
A query will question all these shards (which hold the relevant data) to then collate the results back to you.
I believe you need to review your request and response times over a period of time and if they suffice for your and your business needs then 5 shards just might be the right number for you.
You should always consider that shards do consume resources and the more you have the more resources are used. So considering do you need replica = 2?
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.