Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/24/12 2:26 AM, Andreas Ericsson wrote: > I'd have to send you the new Nagios code and get Sven to help me patch > mod_gearman to avoid using threads, but if you want to give it a shot, > I'm sure we could have you up and running in notime. We're going with something else in the short term, probably updating our obsess commands to send data to multiple local servers, and pushing freshness checking off the master, onto those local failover nodes. We finally reached the point where freshness checking on the master in Nevada, if there was a problem in Europe, could actually make the master fall far enough behind in processing passive checks.. causing cascading failures. Since we have a corporate directive to "Go Linux" we're just taking this as one among *many* reasons to accelerate our migration plan. -- Mike Lindsey -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/24/12 13:06, Andreas Ericsson wrote: > On 02/24/2012 11:37 AM, Sven Nierlein wrote: >> On 2/24/12 11:26, Andreas Ericsson wrote: >>> On 02/23/2012 07:50 PM, Sven Nierlein wrote: On 2/23/12 19:33, Mike Lindsey wrote: > Turns out that's the problem. I've rebuilt from source and it > loads, now to get our package maintainer to rebuild the package. > And to figure out why mod_gearman_worker's children keep > segfaulting. Seems to be freebsd related. A colleague could reproduce that with freebsd 8. >>> Probably caused by fork() in threads. All bsd's are notoriously and >>> retardedly picky about things like that, so programs that work just >>> splendidly on Linux sometimes fail on *BSD. >> Mixing threads and forks is not a good idea on linux too. The NEB module >> part uses threads but no forks. The worker uses only forks but no threads. >> Currently it seems to be a small problem in different behaviour of strdup(). >> > But Nagios uses fork() aplenty after lots of threads have been created. I > know this has been a source of problem many a time before. Thats true, but the Mod-Gearman worker does not share code with Nagios. Anyway, the hint was quite helpful and i got the worker and the neb module running on Freebsd 8 with Gearman 0.25. https://github.com/sni/mod_gearman Thanks for reporting this issue and thanks Andreas for your help :-) Bye, Sven -- Sven Nierlein sven.nierl...@consol.de ConSol* GmbH http://www.consol.de Franziskanerstrasse 38Tel.:089/45841-439 81669 MuenchenFax.:089/45841-111 -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 02/24/2012 11:37 AM, Sven Nierlein wrote: > On 2/24/12 11:26, Andreas Ericsson wrote: >> On 02/23/2012 07:50 PM, Sven Nierlein wrote: >>> On 2/23/12 19:33, Mike Lindsey wrote: Turns out that's the problem. I've rebuilt from source and it loads, now to get our package maintainer to rebuild the package. And to figure out why mod_gearman_worker's children keep segfaulting. >>> Seems to be freebsd related. A colleague could reproduce that with >>> freebsd 8. >>> >> Probably caused by fork() in threads. All bsd's are notoriously and >> retardedly picky about things like that, so programs that work just >> splendidly on Linux sometimes fail on *BSD. > > Mixing threads and forks is not a good idea on linux too. The NEB module > part uses threads but no forks. The worker uses only forks but no threads. > Currently it seems to be a small problem in different behaviour of strdup(). > But Nagios uses fork() aplenty after lots of threads have been created. I know this has been a source of problem many a time before. -- Andreas Ericsson andreas.erics...@op5.se OP5 AB www.op5.se Tel: +46 8-230225 Fax: +46 8-230231 Considering the successes of the wars on alcohol, poverty, drugs and terror, I think we should give some serious thought to declaring war on peace. -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/24/12 11:26, Andreas Ericsson wrote: > On 02/23/2012 07:50 PM, Sven Nierlein wrote: >> On 2/23/12 19:33, Mike Lindsey wrote: >>> Turns out that's the problem. I've rebuilt from source and it >>> loads, now to get our package maintainer to rebuild the package. >>> And to figure out why mod_gearman_worker's children keep >>> segfaulting. >> Seems to be freebsd related. A colleague could reproduce that with >> freebsd 8. >> > Probably caused by fork() in threads. All bsd's are notoriously and > retardedly picky about things like that, so programs that work just > splendidly on Linux sometimes fail on *BSD. Mixing threads and forks is not a good idea on linux too. The NEB module part uses threads but no forks. The worker uses only forks but no threads. Currently it seems to be a small problem in different behaviour of strdup(). > I'd have to send you the new Nagios code and get Sven to help me patch > mod_gearman to avoid using threads, but if you want to give it a shot, > I'm sure we could have you up and running in notime. The problem resides in the worker, although i did not test the neb module yet. Let me first get the worker running on freebsd and then let me try the neb part. Sven -- Sven Nierlein sven.nierl...@consol.de ConSol* GmbH http://www.consol.de Franziskanerstrasse 38Tel.:089/45841-439 81669 MuenchenFax.:089/45841-111 -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 02/23/2012 07:50 PM, Sven Nierlein wrote: > On 2/23/12 19:33, Mike Lindsey wrote: >> Turns out that's the problem. I've rebuilt from source and it >> loads, now to get our package maintainer to rebuild the package. >> And to figure out why mod_gearman_worker's children keep >> segfaulting. > > Seems to be freebsd related. A colleague could reproduce that with > freebsd 8. > Probably caused by fork() in threads. All bsd's are notoriously and retardedly picky about things like that, so programs that work just splendidly on Linux sometimes fail on *BSD. One solution (apart from moving your monitoring infrastructure to Linux) would be to move away from the threaded Nagios and try the up-and-coming Nagios 3.4. It's been running just fine in our lab the past 4 days, although not with mod_gearman enabled. I'd have to send you the new Nagios code and get Sven to help me patch mod_gearman to avoid using threads, but if you want to give it a shot, I'm sure we could have you up and running in notime. -- Andreas Ericsson andreas.erics...@op5.se OP5 AB www.op5.se Tel: +46 8-230225 Fax: +46 8-230231 Considering the successes of the wars on alcohol, poverty, drugs and terror, I think we should give some serious thought to declaring war on peace. -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/23/12 10:50 AM, Sven Nierlein wrote: > On 2/23/12 19:33, Mike Lindsey wrote: >> Turns out that's the problem. I've rebuilt from source and it loads, >> now to get our package maintainer to rebuild the package. And to >> figure out why mod_gearman_worker's children keep segfaulting. > > Seems to be freebsd related. A colleague could reproduce that with > freebsd 8. Any advice short of rebuilding my entire infrastructure? -- Mike Lindsey -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/23/12 19:33, Mike Lindsey wrote: > Turns out that's the problem. I've rebuilt from source and it loads, now to > get our package maintainer to rebuild the package. And to figure out why > mod_gearman_worker's children keep segfaulting. Seems to be freebsd related. A colleague could reproduce that with freebsd 8. -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
On 2/23/12 2:16 AM, Sven Nierlein wrote: > Hi Mike, > > Please don't hijack other threads. Apologies. Unintentional thread header jacking. > Make sure you have eventbroker handling compiled in. > (--enable-event-broker). > Also consider using the latest stable 3.2.3 which has been > successfully tested with > Mod-Gearman. I never tried the 3.3.1. Turns out that's the problem. I've rebuilt from source and it loads, now to get our package maintainer to rebuild the package. And to figure out why mod_gearman_worker's children keep segfaulting. It *looks* like gearman works fine with 3.3.1. At the very least I see jobs going into the queue. -- Mike Lindsey -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null
Re: [Nagios-users] Nagios 3.3.1, event brokers, and debug.
Hi Mike, Please don't hijack other threads. On 2/23/12 4:31, Mike Lindsey wrote: > I'm trying to test out mod_gearman, but I don't see any message about > the event broker loading in the main logfile, and enabling debug logging > just results in a blank debug log file. > > From nagios.cfg: > debug_file=/usr/local/nagios/var/nagios.debug > debug_level=66 > debug_verbosity=2 > max_debug_file_size=1000 > event_broker_options=-1 > broker_module=/usr/local/nagios/lib/mod_gearman.o > config=/usr/local/nagios/etc/gearman.cfg > > From nagios.log: > 1329961651] Successfully shutdown... (PID=79938) > [1329961654] Nagios 3.3.1 starting... (PID=81413) > [1329961654] Local time is Wed Feb 22 17:47:34 PST 2012 > [1329961654] LOG VERSION: 2.0 > [1329961655] Finished daemonizing... (New PID=81414) > > nagios.debug is empty. > > Any advice? Make sure you have eventbroker handling compiled in. (--enable-event-broker). Also consider using the latest stable 3.2.3 which has been successfully tested with Mod-Gearman. I never tried the 3.3.1. Sven -- Virtualization & Cloud Management Using Capacity Planning Cloud computing makes use of virtualization - but cloud computing also focuses on allowing computing to be delivered as a service. http://www.accelacomm.com/jaw/sfnl/114/51521223/ ___ Nagios-users mailing list Nagios-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nagios-users ::: Please include Nagios version, plugin version (-v) and OS when reporting any issue. ::: Messages without supporting info will risk being sent to /dev/null