getting alert that "Replication Headroom has gone below 1 days(avg/sec)". when checked OPS manager for replication lag, max lag I can see that is of 16 seconds. Oplog size has been set to 50gigs. also when checked rs.status(), there is no difference in time for "optimeDate" and "lastHeartbeat". I'm not understanding how to resolve this alert. Also is there any way I can sync secondary and reduce replication lag to zero. Please suggest. Concept explaining shrinking replication headroom/replication lags with its workaround welcomed. Thanks!
1 Answer
"Replication headroom" means "how long time of changes your opLog can store". To get larger "replication headroom" you need to expand your opLog size at least one node, of course it is preferred to have same size opLog at every node.
Reason why this headroom has gone under one day, is that because you have more changes (opLog documents) nowadays, than before.
