That's a FINE question, and there are a million ways to do this. Two quick ideas are:
* Use something like swatch or logwatch against your replication log files and syslog (local4 if I remember correctly) logs.
* Use Big Brother (insert preferred monitoring software <here>), write a simple log parser agent, and then trigger an alert if there is ever unexpected or error output in the log.
I prefer exception-style reporting, so the second option works better for me.
One of the biggest mistakes I see with clients when they are doing replication is not monitoring for a failure in the replication itself. Failures happen, always, and you need to be prepared to catch it. There is nothing worse than failing over to your standby to find out that it has data that is three weeks old. :)
-- Puryear IT, LLC Identity Management, Directory Services, Systems Integration Baton Rouge, LA * 225-706-8414 * http://www.puryear-it.com
"Best Practices for Managing Linux and UNIX Servers" http://www.puryear-it.com/pubs/linux-unix-best-practices
Thierry Lacoste wrote:
Also, be sure to put into place log monitoring for slapd (and possibly slurpd if you are replicating) to watch for errors.
What tool do you suggest for log monitoring?
Regards, Thierry.