I have created Sharded cluster with 3 shards. while balancer was moving the chunk, i executed count query on sharded collection & found that it is giving mismatched count.

I came to know about cleanupOrphaned command which removes invalid documents from the shard & executed on the primary node. It has taken more than 4 days time (i.e. 96 hours) to remove 1.8M records.

First of all why we need to run this command manually, it should have been automatically handled during balancing process. Even after running this command, It takes too much time to remove document. Is there any faster method to clear orphaned file?

Thanks.....