uPortal IRC Logs-2007-05-03

Owned by Apereo Infrastructure
Last updated: Oct 09, 2007Version comment
6 min read

[08:37:32 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal
[09:03:24 EDT(-0400)] * jayshao (n=jayshao@wg-209-hill012.rutgers.edu) has joined ##uportal
[09:34:24 EDT(-0400)] * esm (n=esm@clue.mse.jhu.edu) has joined ##uportal
[10:11:14 EDT(-0400)] * jayshao_ (n=jayshao@wg-209-hill012.rutgers.edu) has joined ##uportal
[10:28:29 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:34:07 EDT(-0400)] * EricDalquist_pid (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[10:34:48 EDT(-0400)] * EricDalquist_pid (n=dalquist@bohemia.doit.wisc.edu) has left ##uportal
[10:58:26 EDT(-0400)] <EricDalquist> I really don't like the eclipse svn plugin
[11:01:26 EDT(-0400)] <esm> ?
[11:01:35 EDT(-0400)] <esm> if you limit yourself to commits (smile)
[11:01:38 EDT(-0400)] <EricDalquist> when you share a project
[11:01:43 EDT(-0400)] <EricDalquist> it doesn't create the trunk/tags/branches
[11:01:55 EDT(-0400)] <EricDalquist> it just dumps the project into the exact dir you specify
[11:02:02 EDT(-0400)] <EricDalquist> makes sense to a point
[11:02:03 EDT(-0400)] <esm> ah i'm not familair with the project sharing
[11:02:06 EDT(-0400)] <EricDalquist> but annoying
[11:02:09 EDT(-0400)] <esm> yeah
[11:02:12 EDT(-0400)] <EricDalquist> just the inital commit
[11:02:19 EDT(-0400)] <EricDalquist> I figured it would be smart enough to add /trunk
[11:02:41 EDT(-0400)] <esm> i guess they don't want to presume to know what the user wants
[11:02:47 EDT(-0400)] <esm> (aka microsoft)
[11:02:52 EDT(-0400)] <EricDalquist> yeah
[11:06:50 EDT(-0400)] * jayshao (n=jayshao@wg-209-hill012.rutgers.edu) has joined ##uportal
[11:16:20 EDT(-0400)] * Colin_ (n=colin@142.150.154.101) has joined ##uportal
[11:20:49 EDT(-0400)] <esm> colinclark: Hi Colin, its Elliot
[11:20:56 EDT(-0400)] <esm> nice to see you in the chat room (smile)
[11:21:00 EDT(-0400)] <colinclark> Hey Elliot.
[11:21:26 EDT(-0400)] <colinclark> I'm still trying to get Trillian, my client on windows, working properly for IRC.
[11:21:32 EDT(-0400)] <colinclark> Nice to see you, too.
[11:21:35 EDT(-0400)] * esm nods
[11:21:52 EDT(-0400)] <esm> good luck, we have some trillian users here but i don't know any of them personally
[11:21:58 EDT(-0400)] <EricDalquist> I used to use it
[11:22:10 EDT(-0400)] <EricDalquist> then pidgin (used to be gaim) 2 came out
[11:22:20 EDT(-0400)] <EricDalquist> and it does everything I used trillian for
[11:22:58 EDT(-0400)] <esm> ok so one, hopefully last thing to take care of: OOM error when assembling large ( >20 Mb) wars
[11:23:16 EDT(-0400)] <EricDalquist> really?
[11:23:25 EDT(-0400)] <EricDalquist> in the stream based WAR code I wrote?
[11:23:38 EDT(-0400)] <esm> yeah, but I've modified it so this may be my fault
[11:23:46 EDT(-0400)] <EricDalquist> (smile)
[11:24:00 EDT(-0400)] <colinclark> Thanks, Eric. I'll have to check it out.
[11:25:55 EDT(-0400)] <colinclark> Quick question. One of our developers is just in the process of building uPortal 3 and mentioned that a unit test is failing. AssemblingWARArchiverTest.testEarAssembly(), apparently.
[11:26:16 EDT(-0400)] <colinclark> Should we worry about it, or just skip the tests while building?
[11:26:30 EDT(-0400)] <colinclark> We all do test-first development here, so they get nervous skipping tests. (smile)
[11:26:39 EDT(-0400)] <EricDalquist> um, check with esm
[11:26:51 EDT(-0400)] <EricDalquist> he is changing much of that code right now
[11:27:24 EDT(-0400)] <colinclark> k, thanks.
[11:28:09 EDT(-0400)] <esm> colinclark: those tests shouldn't fail
[11:28:28 EDT(-0400)] <esm> can you use http://rafb.net/paste and paste the output
[11:28:52 EDT(-0400)] <esm> including the output from uportal3-components/target/surefire-reports/<the class test name that failed>.txt
[11:29:29 EDT(-0400)] <colinclark> Sure.
[11:29:48 EDT(-0400)] <esm> while those tests shoudln't fail, the assembly in trunk is broken
[11:30:02 EDT(-0400)] <esm> so I'm working on a fix
[11:30:19 EDT(-0400)] <esm> but still, those tests shouldnt' fail so I'll take a look
[11:30:27 EDT(-0400)] <colinclark> Thanks!
[11:30:34 EDT(-0400)] <esm> np
[11:38:51 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[11:39:02 EDT(-0400)] <EricDalquist> time to try pidgin for irc
[11:41:18 EDT(-0400)] <esm> (smile)
[11:42:03 EDT(-0400)] <EricDalquist> intersting
[11:42:14 EDT(-0400)] <EricDalquist> I can add the ##uportal chat room to my buddy list
[11:42:17 EDT(-0400)] <EricDalquist> that is really cool
[11:42:19 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has left ##uportal
[11:45:47 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[11:45:52 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has left ##uportal
[12:02:23 EDT(-0400)] <colinclark> esm: Here's the test failure report. http://rafb.net/p/0byEfj48.html
[12:06:30 EDT(-0400)] <esm> colinclark: thanks I will take a look!
[12:07:17 EDT(-0400)] <esm> colinclark: one more thing - what is the output of java -version on that machine?
[12:07:24 EDT(-0400)] <esm> i suspect the problem is spaces in directory names
[12:09:50 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[12:10:09 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has left ##uportal
[12:10:29 EDT(-0400)] * EricDalquist (n=dalquist@bohemia.doit.wisc.edu) has joined ##uportal
[12:31:31 EDT(-0400)] <esm> colinclark: try this - http://rafb.net/p/N2xrfe66.html
[12:31:36 EDT(-0400)] <esm> let me know if you ahve trouble applying the patch
[12:32:16 EDT(-0400)] <esm> i'm away in meetings this afternoon but I will check back here and read the logs as well
[13:04:47 EDT(-0400)] * jayshao (n=jayshao@jshao.rutgers.edu) has joined ##uportal
[13:30:20 EDT(-0400)] * colinclark (n=colincla@142.150.154.101) has joined ##uportal
[13:34:50 EDT(-0400)] <colinclark> esm: thanks for the patch!
[15:12:35 EDT(-0400)] <EricDalquist> hey esm is there a way to find out the first rev of a file in the repo without dumping the entire log for the file?
[15:27:29 EDT(-0400)] <esm> EricDalquist: i don't think so - i don't know of any unfortunatly
[15:27:36 EDT(-0400)] <EricDalquist> ok
[15:29:00 EDT(-0400)] <esm> so IOUtils seems to be blowing up
[15:29:12 EDT(-0400)] <EricDalquist> with memory issues?
[15:29:12 EDT(-0400)] <esm> but I have to do debugging on a machine at home since I have other stuff to do at work
[15:29:13 EDT(-0400)] <EricDalquist> which command
[15:29:22 EDT(-0400)] <EricDalquist> er api call
[15:29:54 EDT(-0400)] <esm> one sec
[15:29:55 EDT(-0400)] <esm> o
[15:30:01 EDT(-0400)] <EricDalquist> ooooo
[15:30:03 EDT(-0400)] <EricDalquist> I see the bug
[15:30:04 EDT(-0400)] <esm> IOUtils.copy( inputstream, outputstream)
[15:30:06 EDT(-0400)] <EricDalquist> yeah
[15:30:11 EDT(-0400)] <EricDalquist> oh
[15:30:17 EDT(-0400)] <EricDalquist> I think I see a different possible bug
[15:30:21 EDT(-0400)] <EricDalquist> I'll look at that too
[15:30:29 EDT(-0400)] <esm> what is the other bug
[15:30:42 EDT(-0400)] <EricDalquist> it wasn't
[15:30:46 EDT(-0400)] <EricDalquist> I looked at the code wrong
[15:30:59 EDT(-0400)] <esm> oh ok i'm going to send a paste one sec
[15:31:10 EDT(-0400)] <EricDalquist> k
[15:31:57 EDT(-0400)] <esm> hm
[15:32:21 EDT(-0400)] <esm> http://rafb.net/paste/pQuiwv96.html
[15:32:38 EDT(-0400)] <esm> compare line 7 to line 9
[15:32:39 EDT(-0400)] <EricDalquist> 404
[15:32:43 EDT(-0400)] <esm> er
[15:32:53 EDT(-0400)] <esm> http://rafb.net/paste/p/pQuiwv96.html
[15:33:02 EDT(-0400)] <esm> wtf
[15:33:13 EDT(-0400)] <EricDalquist> 404 :/
[15:33:58 EDT(-0400)] <esm> http://rafb.net/p/pQuiwv96.html thre
[15:34:00 EDT(-0400)] <esm> there
[15:34:06 EDT(-0400)] <esm> but i have an idea where it may be coming from
[15:34:43 EDT(-0400)] <esm> Used memory: 35417992 Free memory: 16818296 Total memory 52236288
[15:34:49 EDT(-0400)] <esm> Used memory: 60320048 Free memory: 25474768 Total memory 85794816
[15:34:58 EDT(-0400)] <EricDalquist> uhg
[15:39:24 EDT(-0400)] <EricDalquist> what do you think the culprit may be?
[15:40:45 EDT(-0400)] <esm> at first I thought it was actually inside IOUtils, which really seemed unlikely to me from the giddyup. I think there are large buffers being created somewhere, obviously on accident. Like, new ByteArrayInputStream( byteArrayOfTheEntireWar ) or something like that.... I'm not sure I just added more debugging
[15:40:56 EDT(-0400)] <esm> outside of IOUtils
[15:41:00 EDT(-0400)] <esm> as that is battletested
[15:41:54 EDT(-0400)] <EricDalquist> I wonder if it is as simple as calling .flush() in the while loop that iterates over the JAR entries
[15:42:08 EDT(-0400)] <EricDalquist> I just peaked in IOUtils and I can't see how that would leak
[15:42:30 EDT(-0400)] <esm> yeah
[15:42:32 EDT(-0400)] <EricDalquist> they create one 4K byte array, scoped to the copy method call, and that's it
[15:42:44 EDT(-0400)] <esm> yup...
[15:42:50 EDT(-0400)] <esm> i'll move the flush up
[16:06:15 EDT(-0400)] * colinclark (n=colincla@142.150.154.101) has joined ##uportal
[16:13:54 EDT(-0400)] <esm> colinclark: i read the scrollback - did that patch work for you?
[16:14:12 EDT(-0400)] <esm> it wasn't clear if you were thanking me for providing a working patch, or for just providing a patch (smile)
[16:14:34 EDT(-0400)] <EricDalquist> lol
[16:15:01 EDT(-0400)] <esm> hehe (smile)
[16:15:07 EDT(-0400)] <esm> if the patch worked I'll commit it
[16:15:26 EDT(-0400)] <esm> come to think of it, now that I have my windows virtual machine handy i guess I could test it myself
[16:15:33 EDT(-0400)] <esm> create a snapshot image with tomcat and uportal on it
[16:38:30 EDT(-0400)] <EricDalquist> yay I found the first rev of the good load into the uP2 folder
[16:38:42 EDT(-0400)] <EricDalquist> this is looking to be some serious svnfoo
[16:42:58 EDT(-0400)] <esm> is this concerning the corrupt revision that ohlo had a problem with?
[16:43:17 EDT(-0400)] <EricDalquist> well I'll probably resolve that too
[16:43:42 EDT(-0400)] <EricDalquist> but it looks like fisheye isn't happy with all the loads we've done into that cvs staging dir
[16:43:51 EDT(-0400)] <esm> svn ninja kind of rolls off the tounge
[16:43:56 EDT(-0400)] <EricDalquist> so my plan is to do a svnadmin dump
[16:44:11 EDT(-0400)] <EricDalquist> use dumpfilter to drop the cvsstaging and up2 folders
[16:44:15 EDT(-0400)] <EricDalquist> import that
[16:44:29 EDT(-0400)] <esm> yeah
[16:44:36 EDT(-0400)] <EricDalquist> do a svn dump from rev 28654 to now
[16:44:49 EDT(-0400)] <EricDalquist> use dumpfilter to filter eveything but the up2 folder
[16:44:51 EDT(-0400)] <EricDalquist> then import that
[16:44:58 EDT(-0400)] <EricDalquist> and we should end up with a 'nice' svn repo again
[16:46:52 EDT(-0400)] <EricDalquist> so yeah
[16:46:55 EDT(-0400)] <EricDalquist> svn ninja action
[16:47:33 EDT(-0400)] <esm> yeah i shudder at dumpfilter
[16:47:39 EDT(-0400)] <esm> i have only used it once
[16:48:12 EDT(-0400)] <EricDalquist> well I have a local repo that I'm populating right now as a clone of the jasig repo
[16:48:12 EDT(-0400)] <esm> ok i'm leaving early talk to you all soon
[16:48:16 EDT(-0400)] <esm> cool
[16:48:19 EDT(-0400)] <EricDalquist> and eveyrthing will be done there first
[16:48:20 EDT(-0400)] <EricDalquist> multiple times
[16:48:22 EDT(-0400)] <EricDalquist> (smile)
[16:48:23 EDT(-0400)] <EricDalquist> later
[16:48:27 EDT(-0400)] <esm> hehe later
[17:34:34 EDT(-0400)] * colinclark (n=colincla@142.150.154.101) has joined ##uportal
[18:37:22 EDT(-0400)] * colinclark (n=colincla@142.150.154.101) has left ##uportal
[23:29:17 EDT(-0400)] * jayshao (n=jayshao@ool-44c59034.dyn.optonline.net) has joined ##uportal