I'm fairly new to ES so this may be a simple question. If I have a "cluster" of one node that has two indices on the node, is there any benefit to having number_of_shards greater than one?
As I understand it this will split the index across two shards, but won't provide any resiliency in the way a replica will. Is that correct?
The number of replicas is controlled by the number_of_replicas setting. Defaults to 1 which means one copy of each primary.
But if you have only one node, replicas won't be allocated.
There is. Like if you are indexing a lot of data per second, writing in 2 parallel "index" (here shards) could help for example. Or if you have more than 20-50 gb of index size, keeping the size per shard around 20-50 gb will make sense.
But with a "basic", "small" use case, one shard id perfectly fine.
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.