We are running MongoDB Sharded Cluster in our production environment. We set 1 GB for chunk size. After running about 7 months one collection has triggered rebalancing. We found that this balancing due to chunk size of that collection changed from 1 GB to 400 MB. Unfortunately, balancing on high traffic made the latency of our application increasing. Can we know what is the reason of suddenly changed of chunk size?
current chunk size of affected collection
chunk size configuration