Robust Error Monitoring?

manage 1000s of separate web apps, need to monitor them

[Jackson Hole]

Hi there, friends!

I'm tasked with managing 1000s of small web-based applications, written in ColdFusion. Each of these, occasionally, throws an error. These errors are primarily caused by database issues (database is down, table is locked, query timed out) but sometimes they are caused by poorly written applications that don't validate user data well, and a slew of other things that I won't get into here.

Currently, any unhandled exception in any of these apps sends an email to me and my team. We monitor them, notice spikes/trends, and fix accordingly.

I would like to come up with a better way to monitor these application errors. I'm hopeful that there is some open source software/component that I can integrate with, but I'm also open to coding something myself.

I'm also interested in monitoring the health of the servers these apps live on - sort of a separate task, but related, for this I'd definitely want something open-source to just install and configure.

Any ideas on either?

Thanks!

-Jackson

[encyclo]

For ColdFusion, I believe the choice is somewhat limited - you would probably need to code something yourself. As for the servers, what platform do you use?

[pageoneresults]

Hmmm, according to this recent topic, you shouldn't have any errors. ;)

Tracking of 400 and 500 Errors Part 2
Are you being Proactive or Reactive
[webmasterworld.com...]

[Jackson Hole]

Thanks for the replies, that other thread is interesting. I'll have to just write perfect code from here on out. :)

We're on Red Hat, Solaris, Oracle, and Apache.

Anyone have any suggestions on what would be a good architecture for something custom I might write myself? Easiest is to log to a database table, but then I'd need logic to log to a file or email if the database call fails. I guess it's not that complicated...maybe I should just start coding.