True Confessions - I'm finally implementing OpenLog...
I have a confession to make... when it comes to error handling in code, I'm very lax.
There... I've said it. The guilt of listening to multiple Lotusphere speakers telling me I am a bad coder got to be too much...
Seriously, I've wanted to implement OpenLog for quite some time. But it's one of those Covey "important but not urgent" activities that never quite seemed to get my time when it should have. Today I decided to shelve the "urgent but not important" items for a couple of hours and set up OpenLog on our development server. In the matter of about an hour, I had the database configured and two applications writing event messages out to it. I have a proposal out to the developers and admins to, by the end of this month, have the OpenLog database moved to production, a replica placed on all the production application servers, and about a dozen of my applications configured to send at least start/stop times to OpenLog.
These particular jobs are ones that process domlog.nsf each night, looking for different items from the day before. I've had agent timeout issues before, but I've been flying somewhat blind in the overall timing and configuration of the agent scheduling. Now with OpenLog, I'll have a much better idea as to how long each is taking, as well as a heads-up if there's timeout issues (before the user has to alert me to the problem). I also will be motivated to start putting basic error handling in all my agents, as well as retrofitting agents in existing applications.
Putting that in place almost feels like a cleansing of the soul... :)
Thanks again, Julian!