2 Replies Latest reply on May 17, 2012 9:29 AM by DickenWinyard

    Filemaker Server 10 Crashes every 24 hours

    DickenWinyard

      Looking for some insight/good-guesses here:

       

      Server: Filemaker Server 10

      OS: Mac OS X 10.5.8

      Hardware: Mac Mini Core 2 Duo, 2GB RAM

       

      This machine has been running in the above configuration, reliably, for *years*.

       

      For the last 2-3 weeks, it has crashed every single day, somewhere between midnight and 2AM. The exact time varies every day. The range of crash times precisely is from 0.06AM through to 1.52 AM, with no apparent pattern otherwise.

       

      NO Filemaker schedules or backups run during these times.

      None of the user cron-jobs setup on the server run at this time.

       

      I have pored over the server console logs, filemaker logs, etc., to try and find some common process that is running at the same time as the crash, but without luck.

       

      The thread that crashes in fmserver is *mostly* (80%):

       

      Thread 0 Crashed:

      0 libSystem.B.dylib 0x92015f19 _longjmp + 33

       

      But sometimes I see (maybe 10%):

       

      Thread 0 Crashed:

      0 libSystem.B.dylib 0x92058d52 __kill + 10

      1 libSystem.B.dylib 0x920cb242 raise + 26

      2 libSystem.B.dylib 0x920d7681 abort + 73

      3 libstdc++.6.dylib 0x962e7005 0x9629f000 + 294917

      4 libstdc++.6.dylib 0x962e510c __gxx_personality_v0 + 1108

      5 libstdc++.6.dylib 0x962e514b std::terminate() + 29

      6 libstdc++.6.dylib 0x962e5261 __cxa_throw + 101

      7 ...ilemaker.omniorb4.framework 0x0117007a omni_condition::~omni_condition() + 78

      8 libSystem.B.dylib 0x92013c9c __cxa_finalize + 241

      9 libSystem.B.dylib 0x92013b90 exit + 33

      10 com.scodigo.PHP 0x17831a6e _zend_bailout + 72

      11 com.scodigo.PHP 0x178179eb php_error_cb + 1042

      12 com.scodigo.PHP 0x17a2929a soap_error_handler + 856

      13 com.scodigo.PHP 0x178323ed zend_error + 1228

      14 com.scodigo.PHP 0x17821548 zend_timeout + 137

      15 libSystem.B.dylib 0x920571cb _sigtramp + 43

      16 ??? 0xffffffff 0 + 4294967295

      17 libSystem.B.dylib 0x92062449 pthread_cond_wait + 48

      18 ...filemaker.support.framework 0x015dda6f Draco::Semaphore::P(Draco::Thread*) + 159

      19 ...filemaker.support.framework 0x015de585 Draco::MsgQ::GetMsg() + 69

      20 fmserverd 0x0000ee2f ServerApp::OnRun() + 45

      21 com.filemaker.FMS.framework 0x006f64d5 FMS::AppBase::StartService() + 85

      22 fmserverd 0x0000e6c4 main + 470

      23 fmserverd 0x00002d56 _start + 216

      24 fmserverd 0x00002c7d start + 41

       

      Any idea on what might be causing this before I give up and rebuild the system from scratch?

        • 1. Re: Filemaker Server 10 Crashes every 24 hours
          ch0c0halic

          Dicken may be running on logs,

           

          2   libSystem.B.dylib                       0x920d7681 abort + 73

          0   libSystem.B.dylib                       0x92058d52 __kill + 10

          Indicates a library operation hung and tried to abort and then was automatically 'killed - force quit'.

          This library may be part of the following plugin, not necessarily part of the OS.

           

          10  com.scodigo.PHP                         0x17831a6e zendbailout + 72

          Scodigo is the SmartPill PHP plugin manufacturer. Indicates a plugin may be having a problem.

           

          16  ???                                     0xffffffff 0 + 4294967295

          Indicates something isn't registered correctly with the OS. Probably corruption.

           

          The rest is the FMS stuff. I'd guess from its position on the list that its what started the 'thread' but not what caused the crash.

           

          My suggestion is to update the SmartPill plugin and see fi that fixes it.

          • 2. Re: Filemaker Server 10 Crashes every 24 hours
            DickenWinyard

            Helllo Ch0c0halic, and thank you for the pointers.

             

            I have replaced the Smartpill Plug-in to ensure it wasn't corrupted. I haven't upgraded it to the latest version, as we use version 1.5 on our other servers (without problem), and I need things consistent.

             

            The problem has not gone away. Disk Utility finds no problem with the files, and disk-cloning software reports no corruption either when it clones the disk.

             

            I altered the run times of some scripts, and the 'regular time of day' during which the crashes occur have changed - although they are still nowhere near any of the scripts. However, I'm guessing now that one of the scheduled scripts is getting into an endless loop or timing out on something, and continuing to investigate along those lines....