I'm currently working on setting up AWS Elasticsearch infrastructure in my organisation.
During bulk indexing operation (write operation), I have observed that my data nodes gets upscaled to 6 nodes and once the bulk indexing gets completed, it's again downscaled back to 3 data nodes.
On AWS Elasticsearch dashboard, I've already defined number of data nodes as 3 (default).
Number of replicas is set to 1 (post indexing) and number of shards is set to 7.
My question is more related to upscale and downscale process of nodes. How AWS is managing this process during bulk indexing process? How these scale-ups are happening?
AWS Elasticsearch does not auto scale for indexing or any operation in general.
If you observe data nodes doubling and then coming back to original count, it could be that you have also made a configuration change and AWS Elasticsearch has triggered a blue/green deployment.