| jodh | xnox/stgraber: http://people.canonical.com/~jhunt/upstart/gui/upstart-monitor.py | 12:05 |
|---|---|---|
| xnox | =) nice. Will try that in a bit. | 15:27 |
| xnox | jodh: are you ok if I cherry-pick and upload the reboot command commit into ubuntu raring? | 15:28 |
| * xnox notes to procenv compate upstart and normal sessions. | 15:29 | |
| xnox | s/compate/compare/ | 15:29 |
| jodh | xnox: as long as it's tested, sure :) | 15:29 |
| xnox | it totally works on nexus tablets. | 15:29 |
| xnox | no evident breakage on !armhf | 15:29 |
| pabelanger | Anybody able to point to an example of having upstart 'start on' a service that _may_ not be installed? | 17:34 |
| SpamapS | pabelanger: we've talked about this before. :) | 17:42 |
| SpamapS | haven't we? | 17:42 |
| SpamapS | perhaps that was somebody else | 17:42 |
| SpamapS | pabelanger: anyway, if a service can be on the box and optional, it can also be on another box, right? | 17:44 |
| SpamapS | pabelanger: thus, your program should handle it not being ready yet, in case you have started the two boxes in the wrong order. | 17:44 |
| SpamapS | pabelanger: if, however, its something else, like your service supports the other one (like nmbd supports winbind), then you should use wait-for-state in pre-start... | 17:45 |
| SpamapS | pabelanger: as in 'if [ -e /etc/init/$otherjob.conf ] ; then start wait-for-state WAIT_FOR=otherjob WAITER=$UPSTART_JOB ; fi | 17:46 |
| * xnox likes checking `status $otherjob` as the job configuration file can be in multiple source directories or not on disk at all (in case of e.g. user session upstart and/or stateful re-exec) | 18:02 | |
| SpamapS | xnox: *good* point | 18:16 |
| SpamapS | I believe actually wait-for-state will usually return immediately if the job doesn't exist because it does fail the start | 18:17 |
| pabelanger | SpamapS, great, let me check the documentation bout that not | 18:40 |
| pabelanger | now* | 18:40 |
| soso | hello all | 21:29 |
| soso | is someone from here interested into bugs with init 5 command, halt command and with services which don't running correctly when I am using halt command? | 21:31 |
| soso | in ubuntu 12.04 LTS exactly | 21:32 |
| JanC | soso: I think it's best if you say what your problems are and then hang around until somebody sees it (which might be tomorrow...) | 21:37 |
| xnox | soso: also filing bug reports on launchpad either against upstart (ubuntu package) or upstart (upstream projects) is also an option. | 22:15 |
| xnox | soren: some of the developers are in night timezone already though. | 22:15 |
| xnox | soren: but yeah, go ahead =) what are you experiencing? | 22:15 |
| xnox | we are currently working on improving shutdown. | 22:15 |
| soso | So I am running Ubuntu 12.04 LTS and there is an issue is with init 5 command which switch me to run lvl 5 but this command is not able to kill applications which are not part of OS. I am not sure whether init 5 is using only command kill or also kill -9. I didn't read init 5 script. Anyway, after when I am typing command halt system is freezing (hanging). The only chance how to bring system up is with reset button on my PC. But... when I am po | 22:20 |
| soso | wer-cycle system I have a issue with services, exactly any network services (it looks like problem with services, but I don't know how to check it. I am new on Ubuntu). I am able to get network, but I have realy slow response on internet. When I am rebooting system again all services are going up normaly and everything is working correctly. I can send you some outputs if you will need. | 22:20 |
| soso | if you need* | 22:22 |
| soso | sorry for my English :P but I am sure that you are understand :D | 22:23 |
| xnox | soso: upstart kills jobs that it manages (almost everything in /etc/init/*) but than later on sendsigs /etc/init.d/sendsigs actually kills the rest of things. | 22:24 |
| xnox | but it gets interesting as one cannot kill things that are needed to unmount the root filesystem, so if for example you have networked storaged attached there are configuration where one can get stuck. | 22:25 |
| xnox | (e.g. not being able to shutdown cleanly) | 22:25 |
| xnox | soso: I'd recommend you file a bug against upstart and describe your setup and attach any logs that help (e.g. increase boot / shutdown verbosity) | 22:26 |
| xnox | and hopefully some of upstart developers will be able to investigate further. | 22:26 |
| soso | It is not about NAS or SAN. I am talking about slow response from Ethernet connection | 22:32 |
| xnox | interesting. | 22:36 |
| soso | yea. I am not experienced in Ubuntu or networking. so you can write me anytime that I don't know what I am talking about :D | 22:40 |
| JanC | it might still be useful to report a bug | 22:41 |
| JanC | and add/attach as much info as you can | 22:41 |
| soso | yes, I am still here because I need to know which attached information can be usefull for this case. or from which commands I can get best information before I will report it | 22:43 |
| soso | ok. so ps -ef can be usefull to check which processes are not able to shut down after init 5 | 22:44 |
| soso | any another idea what can be usefull? | 22:45 |
| soso | useful*, sorry | 22:45 |
| xnox | /var/log/syslog /var/log/upstart/* | 22:47 |
| xnox | as well. | 22:47 |
| xnox | nothing else spings to mind out.... | 22:47 |
| JanC | I think if anything else would be useful, it would depend on specific things in those you already named | 22:50 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!