Reindex API is extremely slow

I'm getting a write speed of 680 KB/s per node for the reindex API.

Cluster is a 16 data node cluster with 3 master nodes. Each data node has 1.3 TB gp3 Disk with 3000 IOPS and 125 MBps throughput.

Reindex API is using a batch of 6176 documents (~ 3MB payload size)

Why is it so slow?

Which version of Elasticsearch are you using? How large is the index you are reindexing? How many primary and replica shards does it have? How are you running the reindex operation? Have you tried slicing to improve concurrency? Have you tried with different bulk sizes?

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