Alert: zLagTime num > n

What does it mean? What do I do about it?

Note: The same variable name is currently being used in two places, for Pro2 Replication and for OE Replication. A correction is expected in the next release of ProTop.

Example alert from the ProTop Portal for Pro2:

Or for OE Replication:

What does it mean?

In the first image, this is the age of the oldest record in the given queue and has exceeded the alert threshold. You know this is for Pro2 as all the attributes listed in the last line begin "rq_".

The second image indicates that your OE Replication agent is falling behind the source replication server by the indicated number of seconds. This is for OE Replication monitoring as the red box contains "dragent".

What to do?

Pro2

Investigate using the Pro2 Admin utility. Is there an anomaly?

Look at ProTop RT: Pro2 Queues (the "2" command key) is the queue in a running state?

Is the current queue lag consistent with your Pro2 Trends on the ProTop Portal?

If normal (or a new normal), you can increase the alert threshold in etc/alert.*.cfg to a level you want to know about.

If all is running as expected, but the queue is not keeping up, you may need to do some load balancing in your Pro2 configuration.

OE Replication

Investigate the state of your OE Replication for this database as you normally do.

Look at ProTop RT: Replication Agents (the "r" command key).

Or run this command on the target db:

dsrutil /data/db/<dbname> -C monitor

More often than not, this may simply require restarting the replication agent on the target db server:

dsrutil <dbname> -C restart agent