In-memory storage engine: delay in acknowledgement from primary when data has successfully replicated

Hi @Kim_Hakseon,

In addition to the information that @chris asked about, rs.conf() output and the specific version of MongoDB server would be helpful context.

The initial log snippet you provided has a w:majority write concern, so one possibility is that the delay you are seeing is related to acknowledgement from a majority of data-bearing voting members rather than replication.

10 minutes sounds like a suspiciously round number – perhaps a timeout or delay. More information on your server version & replica set configuration is needed, but by chance do you have a voting delayed secondary or a secondary with significantly different network or disk latency?

Regards,
Stennie

1 Like