Hi I've recently started transitioning a bunch of our java webapps across to using the JRockit JVM inside our test environment. For the most part its looking good, but given JMX isn't enabled on the apps in the bog standard java environment, I can't make any quantifiable argument that it's any better. That's not really a problem, tests so far a suggest that it's certainly not any slower, and presumably the JIT optimisations and similar that kick in over time will see at least some small speed boost.
One thing I love about JRockit is that the Flight Recorder is enabled by default, completely integrated, so I can now finally start to monitor whats taking place inside what has until now been a black box. Pulling out the information is dead simple, but what I get is utterly baffling as a pure sysadmin. I've spent time today reading up on Java's Garbage collection (is it me or does talking about "infant mortality rates" feel a bit wrong?) I've written a few munin plugins to grab the data and graph it. So far I'm recording the number of live threads, the percentage of garbage heap used, percentage of nursery used, the number of optimised functions, the number of JIT created functions, and and the amount of time spent doing garbage collection (in seconds). It's already throwing up interesting information, like a webapp that is doing virtually nothing in test is spending a lot of time doing garbage collection. I'm wondering if there is anyone here that already has monitoring in place and if so what do you monitor, and how is it useful for you? Also if anyone knows of any appropriate primers that might get me up to scratch on this from a sysadmins perspective I'd appreciate it. The amount of information I've been able to dig up about what these items actually mean is fairly minimal. Paul _______________________________________________ Tech mailing list [email protected] http://lopsa.org/cgi-bin/mailman/listinfo/tech This list provided by the League of Professional System Administrators http://lopsa.org/
