Adding new Slave to existing Mongo cause the Timeout in Application

Ops and Admin Replication
1

Hi,

We are having 2.7TB of data on primary and due to some reason secondary went down and it is now completely out of sync. So now whenever it is getting added into cluster my application starts timing out. Also since slave sync from the beginning it takes almost 2 to 3 days to fully in sync with master.

For application we tried to check if there is issue with write concern but it shows w:1.

Is there is any way we can add Replica without application timeout. Because all operations on master starts timing out as soon as we add Replica. Once replica is in sync we don’t see such issues.

MongoDB Version is 4.2 community version

2

We have a similar problem with our cluster.

We’re trying to sync up a replica using a seeded member set to hidden. When the optimeDate caught up, we brought it into service by setting hidden to false, and immediately everything that connected to this replica for reads was timing out. Looking at rs.status(), the optimeDate just froze. We seem to be missing optimeDurableDate and the optimeDurable: {} from rs.status(), and I can’t find anything to force mongodb to sync up.