We have a sharded cluster with 3 shards. The version is: 3.6.13. The users of the prod. database (1,8 TB sized) complain about an issue. System works normally but after sometime it stucks and doesn't respond. During that time we don't see any long running queries. Everything seems normal in the log files. But when we track the network the requests are coming to Mongos, Mongos sends them to mongod but shards don't send back anything in network level. The single qlue that I catch is about oplog entries. When I took a dump backup with oplog, in dump log it has "dumped 7757639 oplog entries" In the dumps I took last week there was only 1 dumped entry. Is there a problem with the replication that causes waits?
When I check the replication status, it seems there is no problem. Thank you for your replies.