Page Comparison

[09:23:08 EDT(-0400)] * esm (n=esm@asdf.dkc.jhu.edu) has joined ##uportal
[09:53:27 EDT(-0400)] * anastasiac (n=anastasi@142.150.154.154) has joined ##uportal
[10:25:46 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[12:05:27 EDT(-0400)] * JASIGLogBot2 (n=PircBot@bohemia.doit.wisc.edu) has joined ##uportal
[12:05:27 EDT(-0400)] JASIGLogBot2 ##uportal is being logged to: http://www.ja-sig.org/wiki/display/UPC/uportal+IRC+Logs
[12:05:46 EDT(-0400)] <EricDalquist> yay testing
[12:23:25 EDT(-0400)] * colinclark (n=atrcwrk2@142.150.154.101) has joined ##uportal
[12:27:57 EDT(-0400)] * colinclark (n=atrcwrk2@142.150.154.101) has left ##uportal
[12:36:38 EDT(-0400)] * JASIGLogBot (i=jasigch@jasigch.Princeton.EDU) has joined ##uportal
[12:36:38 EDT(-0400)] * Topic is 'http://rafb.net/paste/ to paste. Don't forget about our #fluid-work neighbors' set by EricDalquist on 2007-10-02 14:18:09 EDT(-0400)
[12:38:43 EDT(-0400)] <EricDalquist> testing jasig bot
[12:49:28 EDT(-0400)] * colinclar1 (n=atrcwrk2@142.150.154.101) has joined ##uportal
[12:50:47 EDT(-0400)] * colinclar1 (n=atrcwrk2@142.150.154.101) has left ##uportal
[12:56:09 EDT(-0400)] * pberry (n=pberry@waldorf.CSUChico.EDU) has joined ##uportal
[12:56:24 EDT(-0400)] <pberry> oddly enough #uuportal isn't the right channel
[12:56:35 EDT(-0400)] <EricDalquist>
[12:56:55 EDT(-0400)] <pberry> man, perl is just so messed up
[12:56:59 EDT(-0400)] <EricDalquist> yup
[12:57:04 EDT(-0400)] <EricDalquist> so is ruby
[12:57:11 EDT(-0400)] <pberry> ha
[12:57:15 EDT(-0400)] <pberry> well played sir
[12:57:15 EDT(-0400)] <EricDalquist> but ruby does cool stuff like: http://www.fudgie.org/
[12:57:29 EDT(-0400)] <EricDalquist> I have a nice real time opengl display of our portal activity
[12:57:52 EDT(-0400)] <EricDalquist> well ... as real-time as a machine with no graphics accelerator can do
[12:58:03 EDT(-0400)] <pberry> ha, you're using that too
[12:58:03 EDT(-0400)] <pberry> awesome
[12:58:09 EDT(-0400)] <EricDalquist> yeah
[12:58:18 EDT(-0400)] <EricDalquist> I have it all tweaked out for MyUW
[12:58:31 EDT(-0400)] <EricDalquist> counting logins, doing some URL aggregation
[12:58:48 EDT(-0400)] <EricDalquist> this gives me fond memories of Hackers
[13:00:27 EDT(-0400)] <pberry> EricDalquist: can you pastie your config block?
[13:01:37 EDT(-0400)] <EricDalquist> https://mywebspace.wisc.edu/dalquist/web/gl_tail-0.03.rb
[13:01:44 EDT(-0400)] <EricDalquist> ssh keys
[13:01:51 EDT(-0400)] <EricDalquist> so no worry about looking at the whole file
[13:02:56 EDT(-0400)] <pberry> ah, you're parsing apache and not portal.log
[13:03:06 EDT(-0400)] <EricDalquist> yeah
[13:03:21 EDT(-0400)] <EricDalquist> we only log WARN/ERROR to the portal log
[13:03:25 EDT(-0400)] <EricDalquist> so that isn't that interesting
[13:04:39 EDT(-0400)] <pberry> this gives me some good directions to go in
[13:04:40 EDT(-0400)] <pberry> thanks
[13:04:57 EDT(-0400)] <EricDalquist> yup
[13:05:08 EDT(-0400)] <EricDalquist> that config seems to do pretty will on the machine I'm running it on to
[13:05:14 EDT(-0400)] <EricDalquist> usually about 20fps
[13:05:18 EDT(-0400)] <pberry> probably hack it up to monitor portal.log, as we still use that for login/logout
[13:05:25 EDT(-0400)] <EricDalquist> when we get the start/end of class spikes it drops to like 2
[13:06:29 EDT(-0400)] <pberry> on my MBP it clocks in just over 100 FPS
[13:06:51 EDT(-0400)] <pberry> it's kinda fun pointing it at the campus httpd access_log
[13:07:17 EDT(-0400)] <EricDalquist> yeah my MBP at home does that too
[13:07:24 EDT(-0400)] <EricDalquist> but that has a real GFX card in it
[13:07:35 EDT(-0400)] <EricDalquist> the only computers I have at work have onboard intel graphics
[13:07:41 EDT(-0400)] <pberry> this steady stream of image data just constantly flows
[13:08:00 EDT(-0400)] <pberry> but then the fans start to crank up to cool it down...which gets annoying
[13:11:26 EDT(-0400)] <pberry> so, I've got this odd tomcat problem
[13:11:35 EDT(-0400)] <pberry> with one tomcat shutdown taking two instances down
[13:12:05 EDT(-0400)] <pberry> and I can't figure it out
[13:12:35 EDT(-0400)] <EricDalquist> do you have them listening on different shutdown ports?
[13:12:45 EDT(-0400)] <EricDalquist> well first are they on the same box?
[13:12:49 EDT(-0400)] <pberry> I think so /double checks
[13:12:51 EDT(-0400)] <pberry> yeah, same box
[13:12:54 EDT(-0400)] <esm> yeah, how could they listen to the same port?
[13:12:55 EDT(-0400)] <pberry> different jvms
[13:13:04 EDT(-0400)] <EricDalquist> don't know
[13:13:07 EDT(-0400)] <EricDalquist> magic?
[13:13:07 EDT(-0400)] <esm> different interfaces i guess
[13:13:23 EDT(-0400)] <EricDalquist> the shutdown port is on localhost unless you override it
[13:13:49 EDT(-0400)] <esm> grep killall /path/to/shutdown/script
[13:13:58 EDT(-0400)] <pberry> yeah, one is on 2755 and the other is on 8006
[13:14:19 EDT(-0400)] <EricDalquist> so that is really weird
[13:14:40 EDT(-0400)] <EricDalquist> I 'think' the way tomcats shutdown script works is it looks at CATALINA_HOME or some such environment variable
[13:14:49 EDT(-0400)] <EricDalquist> then finds the config for the server based on that path
[13:14:59 EDT(-0400)] <EricDalquist> finds the shutdown port, ip & text from the config file
[13:15:08 EDT(-0400)] <EricDalquist> and contacts that server to shutdown
[13:15:45 EDT(-0400)] <pberry> so, there is no way that one script could contact two different CATALINA_HOME
[13:16:37 EDT(-0400)] <EricDalquist> I wouldn't think so ....
[13:17:04 EDT(-0400)] <pberry> The one that does the killing is run as root
[13:17:15 EDT(-0400)] <pberry> The one that gets killed has it's own uid
[14:28:55 EDT(-0400)] * michelled (n=michelle@142.150.154.114) has joined ##uportal