ruote log_2010-10-23

2010-10-23 00:13:52 utc hassox kk giving it a crack
2010-10-23 00:14:38 utc hassox k not seeing much difference
2010-10-23 00:14:42 utc hassox should I see something?
2010-10-23 00:14:48 utc hassox ACTION already has noisy going
2010-10-23 00:14:50 utc jmettraux it should not hang
2010-10-23 00:14:56 utc jmettraux ok
2010-10-23 00:15:00 utc jmettraux back to step 0
2010-10-23 00:15:08 utc jmettraux how does it hang ? where does it hang ?
2010-10-23 00:15:23 utc hassox on the wait for
2010-10-23 00:15:27 utc jmettraux may I see a faulty noisy ?
2010-10-23 00:16:10 utc hassox yup
2010-10-23 00:16:13 utc hassox getting it for you
2010-10-23 00:17:17 utc hassox https://gist.github.com/48b0038a7b0d2aaafaa4
2010-10-23 00:17:46 utc jmettraux your wait_for is blocking until all the processes are either terminated or emitted an error (not a ${f:error} though)
2010-10-23 00:18:16 utc hassox k
2010-10-23 00:18:34 utc hassox they should all be finished :\
2010-10-23 00:18:40 utc jmettraux "6 20 te * 20101023-bodoseshipi 0 {:wi=>["0_0!!20101023-bodoseshipi", 2]}" tells us that process bodoseshipi terminated ("te")
2010-10-23 00:18:42 utc hassox not sure how to get them to emit an error
2010-10-23 00:18:49 utc hassox oh
2010-10-23 00:19:11 utc hassox there's 6 of those
2010-10-23 00:19:32 utc jmettraux and six terminations
2010-10-23 00:19:42 utc jmettraux my wait_for method is weak
2010-10-23 00:19:51 utc jmettraux (hence my TestLogger try)
2010-10-23 00:20:23 utc jmettraux maybe as a workaround : wait for the last launched process instance
2010-10-23 00:20:24 utc hassox I can't tell if they're all completed
2010-10-23 00:21:03 utc hassox kk trying that
2010-10-23 00:21:46 utc hassox that made it kick through :)
2010-10-23 00:22:35 utc jmettraux ACTION double checking the wait_for(*wfids) tests and code
2010-10-23 00:23:57 utc jmettraux quick note : to emit an error, simply raise, the engine will catch and package that as a process error
2010-10-23 00:24:13 utc hassox oh cools
2010-10-23 00:25:04 utc jmettraux grmpf, I have this http://github.com/jmettraux/ruote/blob/ruote2.1/test/functional/ft_39_wait_for.rb#L61-78 it works
2010-10-23 00:25:13 utc jmettraux I wonder how I could reproduce your issue
2010-10-23 00:26:03 utc hassox by using engine.process_wfids does give me all ids or just ones that are currently still processing?
2010-10-23 00:26:20 utc jmettraux just the ones that are still processing (or in error)
2010-10-23 00:26:32 utc jmettraux when a process terminates it gets wiped out
2010-10-23 00:27:16 utc hassox k
2010-10-23 00:40:18 utc hassox jmettraux: down to 4 red ones!
2010-10-23 00:40:29 utc jmettraux :-)
2010-10-23 00:59:23 utc jmettraux hassox: maybe what happened with your engine.wait_for(*wfids) is that one of your processes already terminated when you called wait_for
2010-10-23 00:59:34 utc hassox that's a very real posibility
2010-10-23 00:59:40 utc hassox doesn't it know that they're done?
2010-10-23 01:03:22 utc jmettraux not really, else it would have to potentially track a growing list of dead wfids
2010-10-23 01:03:52 utc jmettraux it just trusts you
2010-10-23 01:04:43 utc jmettraux of course, for testing purposes, an in-memory list of terminated wfids is ok
2010-10-23 01:09:37 utc hassox i see
2010-10-23 01:09:40 utc hassox ok
2010-10-23 01:09:48 utc hassox so
2010-10-23 01:09:57 utc hassox wait_for doesn't just ignore not found wfids?
2010-10-23 01:11:16 utc jmettraux if you do wfid = engine.launch(pdef); engine.wait_for(wifd);
2010-10-23 01:11:23 utc jmettraux wfid
2010-10-23 01:11:43 utc jmettraux the wfid won't be found, the processes was just launched...
2010-10-23 01:12:15 utc hassox oh right
2010-10-23 01:12:18 utc jmettraux (it happens in another thread or another process, not really instantaneous)
2010-10-23 01:12:22 utc hassox gotcha
2010-10-23 01:12:37 utc hassox threads hurt
2010-10-23 01:12:42 utc hassox my head
2010-10-23 04:08:14 utc hassox I can't test this in my cucumber integration tests :(
2010-10-23 04:11:05 utc jmettraux how can I help ?
2010-10-23 04:14:41 utc hassox I don't know
2010-10-23 04:14:48 utc hassox ACTION is just getting frustrated
2010-10-23 04:15:06 utc hassox can i force the worker to run in the same thread as the cucumber specs?
2010-10-23 04:15:13 utc hassox or will that screw up my app?
2010-10-23 04:15:35 utc jmettraux is that the same transactional_fixtures issue as yesterday ?
2010-10-23 04:20:25 utc hassox kinda
2010-10-23 04:20:29 utc hassox I just can't get it to work
2010-10-23 04:20:31 utc hassox check this
2010-10-23 04:21:31 utc hassox https://gist.github.com/52e999167bc5d82c29f8
2010-10-23 04:21:56 utc hassox but it sometimes runs and sometimes not
2010-10-23 04:22:13 utc jmettraux == sometimes it's green sometimes it's not ?
2010-10-23 04:23:15 utc hassox correct
2010-10-23 04:23:17 utc hassox well
2010-10-23 04:23:24 utc hassox I'm waiting for ti to finish and deliver an email
2010-10-23 04:23:34 utc hassox sometimes it gets delivered.. sometimes not
2010-10-23 04:24:11 utc hassox if I keep running them, eventually I'll get to one failing one, the that will pass, then when they all run again, I get about 12 failures
2010-10-23 04:24:20 utc hassox and it starts again
2010-10-23 04:24:51 utc jmettraux http://gist.github.com/641769
2010-10-23 04:25:05 utc jmettraux (closer to the name of the Then)
2010-10-23 04:25:30 utc jmettraux so many unknowns for me
2010-10-23 04:25:44 utc jmettraux does fail imply not delivered ?
2010-10-23 04:25:46 utc hassox I konw :(
2010-10-23 04:25:49 utc hassox yes
2010-10-23 04:25:55 utc hassox I'm checking the delivered array
2010-10-23 04:26:02 utc jmettraux or are those two things not seemingly related ?
2010-10-23 04:26:39 utc jmettraux are you purging the storage before each test situation ?
2010-10-23 04:26:46 utc hassox yup
2010-10-23 04:28:05 utc jmettraux ok, seen
2010-10-23 04:28:35 utc jmettraux trying to make sense of
2010-10-23 04:28:39 utc jmettraux 13:27 hassox: if I keep running them, eventually I'll get to one failing one, the that will pass, then when they all run again, I get about 12 failures
2010-10-23 04:28:39 utc jmettraux 13:27 hassox: and it starts again
2010-10-23 04:28:53 utc hassox ja I know
2010-10-23 04:28:54 utc hassox me too
2010-10-23 04:30:10 utc jmettraux screenshare maybe
2010-10-23 04:30:26 utc hassox kk
2010-10-23 04:30:35 utc hassox my skype name is dneighman
2010-10-23 04:31:48 utc hassox skype screen share or ichat?
2010-10-23 04:32:43 utc jmettraux which one is more efficient ?
2010-10-23 05:01:17 utc jmettraux puts "WORKER RUNNING ? #{RuoteKit.engine.worker.running}"
2010-10-23 05:02:20 utc jmettraux hassox: puts "WORKER RUNNING ? #{RuoteKit.engine.worker.running}"
2010-10-23 05:11:09 utc jmettraux hassox: http://github.com/jmettraux/ruote/blob/ruote2.1/lib/ruote/id/wfid_generator.rb#L50
2010-10-23 05:17:12 utc jmettraux hassox: I guess we can switch back to the FsStorage now
2010-10-23 05:22:20 utc hassox looking
2010-10-23 05:22:28 utc hassox kk
2010-10-23 05:22:29 utc hassox done
2010-10-23 05:26:39 utc jmettraux is it the same error with FsStorage ? (the error in ruote)
2010-10-23 05:28:02 utc hassox looking
2010-10-23 05:28:04 utc hassox sorry
2010-10-23 05:28:08 utc jmettraux hassox: sorry, could I have the gist of the error again ?
2010-10-23 05:28:09 utc hassox didn't realise you were asking that ;)
2010-10-23 05:28:17 utc hassox ja
2010-10-23 05:28:19 utc hassox 1 sec
2010-10-23 05:29:20 utc hassox nope
2010-10-23 05:29:30 utc hassox in that case, it's not even saving the @employment
2010-10-23 05:29:31 utc hassox :\
2010-10-23 05:30:02 utc hassox bah
2010-10-23 05:30:15 utc jmettraux and no error in your rescue ?
2010-10-23 05:32:13 utc hassox some other issues
2010-10-23 05:32:19 utc hassox the employment is not saving atm
2010-10-23 05:32:31 utc hassox hrm
2010-10-23 05:34:12 utc jmettraux hassox: I can't reproduce the issue here, I was wrong when I said the purge! triggered the @last == nil error
2010-10-23 05:34:19 utc hassox kk
2010-10-23 05:34:30 utc hassox for some reason I can't even get it to fire atm :\
2010-10-23 05:34:51 utc jmettraux ouch
2010-10-23 05:39:02 utc jmettraux hassox: did you reset back the ruote / wfid_generator.rb code ? It should be fine
2010-10-23 05:39:08 utc hassox ja
2010-10-23 05:39:17 utc hassox my db has gone a bit funny
2010-10-23 05:49:50 utc hassox jmettraux: just got my db back under control
2010-10-23 05:49:53 utc hassox same error mate
2010-10-23 05:50:01 utc jmettraux oh great
2010-10-23 05:50:10 utc jmettraux may I see the trace ?
2010-10-23 05:50:37 utc hassox pm'd
2010-10-23 05:51:40 utc jmettraux great
2010-10-23 05:52:12 utc jmettraux wonderful
2010-10-23 05:55:23 utc jmettraux hassox: working on it
2010-10-23 05:55:29 utc hassox thanx so much :)
2010-10-23 05:55:48 utc jmettraux you're welcome m(_ _)m
2010-10-23 06:10:56 utc jmettraux looking good, running a last serie of tests
2010-10-23 06:11:22 utc hassox saweet :D
2010-10-23 06:19:40 utc jmettraux hassox : http://github.com/jmettraux/ruote/commit/3989f44a6bb2f636b15d4458d869099ff32db6c5
2010-10-23 06:20:41 utc hassox :D
2010-10-23 06:20:49 utc hassox ACTION adds the git version to the gemfile
2010-10-23 06:21:28 utc jmettraux should be safe
2010-10-23 06:28:30 utc hassox stand by
2010-10-23 06:28:37 utc hassox bundler is playing up now :(
2010-10-23 06:31:13 utc hassox bundler is done
2010-10-23 06:38:42 utc hassox jmettraux: better
2010-10-23 06:38:46 utc hassox I still have a failure
2010-10-23 06:38:57 utc hassox going through the same process to discover the issue
2010-10-23 06:41:30 utc jmettraux ok
2010-10-23 06:42:26 utc jmettraux hassox: I have to escape soon
2010-10-23 06:42:34 utc hassox ja I've found the issue
2010-10-23 06:42:41 utc hassox could not get 3xx (550)
2010-10-23 06:42:41 utc jmettraux ok
2010-10-23 06:42:52 utc hassox it's not an issue for you
2010-10-23 06:42:56 utc jmettraux ok
2010-10-23 06:42:56 utc hassox it's an smtp issue I think
2010-10-23 06:43:04 utc jmettraux I'll be back in 2-3 hours
2010-10-23 06:43:20 utc jmettraux I leave my IRC session on, feel free to pass me gists / info
2010-10-23 06:43:27 utc hassox cools mate :D
2010-10-23 06:43:29 utc hassox thanx so much
2010-10-23 06:43:52 utc jmettraux no worries
2010-10-23 06:45:30 utc hassox jmettraux: is there a hook to run code in the thread after it threads?
2010-10-23 06:46:02 utc jmettraux the participant thread ?
2010-10-23 06:47:04 utc jmettraux hassox: if it's the participant thread, well there is the consume(workitem) method
2010-10-23 06:47:14 utc hassox nah the worker one
2010-10-23 06:47:27 utc jmettraux if it's the worker thread, then you can hook a listener
2010-10-23 06:47:44 utc hassox how would I do that?
2010-10-23 06:48:04 utc jmettraux you can take that class as a model : http://github.com/jmettraux/ruote/blob/ruote2.1/lib/ruote/log/storage_history.rb
2010-10-23 06:48:43 utc jmettraux copy it, remove all the methods except notify and initialize and you have a "tap" inside of the worker activity
2010-10-23 06:49:03 utc hassox tahnx mate :D
2010-10-23 06:50:52 utc jmettraux you'll receive the messages (msg)
2010-10-23 06:51:07 utc jmettraux it contains all the details about what the worker just did
2010-10-23 06:51:57 utc jmettraux the sister classes in http://github.com/jmettraux/ruote/tree/ruote2.1/lib/ruote/log/ can help as well (it's the logger ones)
2010-10-23 06:52:07 utc jmettraux and the noisy stuff is in pretty.rb IIRC
2010-10-23 06:52:17 utc jmettraux ...escaping...
2010-10-23 06:58:50 utc jmettraux hassox: and here is an example on how to hook the service into the system : http://github.com/jmettraux/ruote/blob/ruote2.1/test/functional/ft_36_storage_history.rb#L23-24
2010-10-23 06:59:16 utc hassox this stuff is gold man
2010-10-23 06:59:22 utc jmettraux I will probably add a engine.worker.add_hook(&block) thing, it could help
2010-10-23 06:59:26 utc jmettraux now have to go...
2010-10-23 06:59:27 utc jmettraux ciao !
2010-10-23 07:00:00 utc hassox k mate
2010-10-23 07:00:03 utc hassox thanx again for your help
2010-10-23 09:09:29 utc jmettraux hassox: back
2010-10-23 09:09:37 utc hassox jmettraux: :D
2010-10-23 09:09:41 utc hassox thanx so much for your help mate
2010-10-23 09:09:57 utc jmettraux you're welcome, how is it going ?
2010-10-23 09:09:57 utc hassox the next issue I had after you made that change was that google had stopped me sending emails ;)
2010-10-23 09:10:07 utc hassox my configs were bad and I was pinging them in my tests
2010-10-23 09:10:08 utc jmettraux lol
2010-10-23 09:10:10 utc hassox hehe
2010-10-23 09:10:11 utc hassox yeah
2010-10-23 09:10:28 utc hassox specs are green for launches now, and my integration tests are functional :D
2010-10-23 09:11:31 utc jmettraux great, if you have some time later in the week, maybe you have suggestions and feedback
2010-10-23 09:24:22 utc hassox I will do yeah
2010-10-23 09:24:33 utc hassox I'm going to consolidate my setup, then I'll run it past you
2010-10-23 09:26:17 utc jmettraux :-)