I've been running an internal C2 application for a while and have noticed what I think are some suspicious database behaviours. Unfortunately I am not able to keep close track of the system to watch as the various pools get used and abused. What I'm wondering is whether it would be possible to add some new items to the output of the StatusGenerator? Off the top of my head it would be nice to know things like: - number of times a database pool has reached the hard maximum - average number of idle/active elements in the pool - how long an item tends to be out of the pool - the state of the pool - min/max/avg/stdev response times (maybe on a per-pipeline basis?) - min/max/avg/stdev requests over a sliding time window - state of the Memory and File stores - the number/rate of info/warn/error/debug statements in the log Anyone who has a thing for dynamically creating SVG graphics might want to make graphs out of this information :) >From my quick look at the Excalibur pooling APIs they don't store any of these metrics, which would make this a bigger change than simply modifying StatusGenerator.java I remember some discussion of what I interpreted to be a profiling sitemap. This is clearly related to this proposal, but would be only one of the things the status generator could take a look at. Any thoughts? Jason Foster --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
