You might want to add the following to main() in gmond.c and try it out apr_signal(SIGTERM, sig_handler);
Try adding this with the other apr_signal() calls in main(). This should instruct SIGTERM to gracefully shutdown in the same way that SIGINT does. Brad >>> On 5/27/2009 at 4:11 PM, in message <[email protected]>, David Birdsong <[email protected]> wrote: > ok, so i just backed up into testing things i have a better grasp on. > > i wrote a super simple python module that conforms to the template for > python modules. metric_cleanup gets called when gmond receives a > SIGINT but not when it receives a SIGTERM. > > the gmond init script installed by default from a fedora rpm will call > 'killproc' which is defined by the distro. killproc defaults to send > a SIGTERM first, wait then send a SIGKILL. > > shouldn't gmond catch a SIGTERM and exit gracefully? > > On Wed, May 27, 2009 at 1:40 PM, David Birdsong > <[email protected]> wrote: >> thanks bernard, that sped me along. >> >> i'm admittedly pretty green to debugger's in general, and i'm not >> quite sure if there's special kung fu in debugging a process's signal >> handler as it, if i understand signals correctly, is installed on a >> different stack. >> >> i: >> - start gmond -d2 in one term >> - attach to the running gmond process >> - set a breakpoint for both apr_terminate and apr_pool_destroy >> - continue >> >> ....gdb will hang while gmond runs it's loop, the non-forked gmond >> window confirms that it's looping periodically collecting metrics and >> sending them off >> >> - send gmond a TERM signal, i get this from gdb: >> >> Program received signal SIGTERM, Terminated. >> [Switching to Thread 140540029593328 (LWP 11780)] >> 0x0000003121ed50d8 in epoll_wait () from /lib64/libc.so.6 >> >> ...maybe that thread is the signal handler? not sure. >> >> it occurred to me that gdb could be interfering with the SIGTERM, so i > added: >> (gdb) handle SIGTERM nostop >> ...but this changed nothing. >> >> i also notice that if i set the breakpoints and left gmond running in >> gdb longer, i'd start hitting apr_pool_destroy.c:716 for every metric >> as it was being sent out(as confirmed by the interactive gmond window) >> >> i could continue for each breakpoint, but this just repeated...i'm not >> sure if i'm actually discovering anything here. >> >> On Wed, May 27, 2009 at 12:41 PM, Bernard Li <[email protected]> wrote: >>> Hi David: >>> >>> On Tue, May 26, 2009 at 6:22 PM, David Birdsong >>> <[email protected]> wrote: >>> >>>> I've been searching around for awhile now, any suggestions on where I >>>> can get apr-debug ...or is it apr-util-debug? I'm on Fedora 8, but >>>> source is fine. >>> >>> Perhaps you are looking for the apr-debuginfo RPM? >>> >>> > http://archives.fedoraproject.org/pub/archive/fedora/linux/releases/8/Everyth > ing/i386/debug/apr-debuginfo-1.2.11-2.i386.rpm >>> >>> Cheers, >>> >>> Bernard >>> >> > > ------------------------------------------------------------------------------ > Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT > is a gathering of tech-side developers & brand creativity professionals. Meet > the minds behind Google Creative Lab, Visual Complexity, Processing, & > iPhoneDevCamp as they present alongside digital heavyweights like Barbarian > Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com > _______________________________________________ > Ganglia-general mailing list > [email protected] > https://lists.sourceforge.net/lists/listinfo/ganglia-general ------------------------------------------------------------------------------ Register Now for Creativity and Technology (CaT), June 3rd, NYC. CaT is a gathering of tech-side developers & brand creativity professionals. Meet the minds behind Google Creative Lab, Visual Complexity, Processing, & iPhoneDevCamp as they present alongside digital heavyweights like Barbarian Group, R/GA, & Big Spaceship. http://p.sf.net/sfu/creativitycat-com _______________________________________________ Ganglia-general mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/ganglia-general
