That depends on your use case. If you have a search use case with a fixed number of indices, and the shard count will not increase further, it may be fine. If you on the other hand are using time-based indices, the larger number of shards generated per time period will cause additional overhead. Also look at measuring query performance at expected production load. having 20 shards may allow you to do more work in parallel when you have few queries, resulting in lower latencies, but this may no longer be the case as the number of concurrent queries grow and a larger number of tasks need to be queued up.
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.