Loading ...
Sorry, an error occurred while loading the content.

Segmentation fault

Expand Messages
  • Tom Cat
    Well I think I ve finally found some more information aboute the seg faults on RHEL 3&4. The ones that segfault never have a LogFile line in the configuration
    Message 1 of 6 , May 8 5:32 AM
    • 0 Attachment
      Well I think I've finally found some more information aboute the seg
      faults on RHEL 3&4. The ones that segfault never have a LogFile line
      in the configuration file (I always specifiy one on the command line).
      The ones that don't segfault have this line, but it gets ignored
      since I'm specifying it on the command line.

      Does this help anyone figure out how to fix the problem?
    • Bradford L. Barrett
      Certain malformed user agent strings will cause a segfaut when you use a MangleAgent level of one (1). Has nothing to do with logfile name or where you
      Message 2 of 6 , May 8 10:48 AM
      • 0 Attachment
        Certain malformed user agent strings will cause a segfaut when you
        use a MangleAgent level of one (1). Has nothing to do with logfile
        name or where you specify it.

        --

        On Mon, 8 May 2006, Tom Cat wrote:

        > Well I think I've finally found some more information aboute the seg
        > faults on RHEL 3&4. The ones that segfault never have a LogFile line
        > in the configuration file (I always specifiy one on the command line).
        > The ones that don't segfault have this line, but it gets ignored
        > since I'm specifying it on the command line.
        >
        > Does this help anyone figure out how to fix the problem?
        >
        >
        >
        >
        >
        >
        >
        > Webalizer homepage: http://www.webalizer.org
        >
        > Yahoo! Groups Links
        >
        >
        >
        >
        >
        >
        >
        --
        Bradford L. Barrett brad@...
        A free electron in a sea of neutrons DoD#1750 KD4NAW

        The only thing Micro$oft has done for society, is make people
        believe that computers are inherently unreliable.
      • Tom Cat
        ... Ok not good folks. I ve just found two cases that invalidated this theory. Going back and repeating the ones that seemed to support this theory, I ve
        Message 3 of 6 , May 8 11:14 AM
        • 0 Attachment
          --- In webalizer@yahoogroups.com, "Tom Cat" <stry_cat@...> wrote:
          >
          > Well I think I've finally found some more information aboute the seg
          > faults on RHEL 3&4. The ones that segfault never have a LogFile line
          > in the configuration file (I always specifiy one on the command line).
          > The ones that don't segfault have this line, but it gets ignored
          > since I'm specifying it on the command line.
          >
          > Does this help anyone figure out how to fix the problem?
          >

          Ok not good folks. I've just found two cases that invalidated this
          theory. Going back and repeating the ones that seemed to support this
          theory, I've found a number of them segfaults about 25% of the time
          but works just fine about 75%. Just my luck that I didn't get a seg
          fault. *sigh*
          Any ideas?
        • Tom Cat
          ... Thanks for the answer. I tried setting MangleAgent to 0 and also to 5. I m still getting the segfault. :( Thinking it could be something in the log, I
          Message 4 of 6 , May 10 7:50 AM
          • 0 Attachment
            --- In webalizer@yahoogroups.com, "Bradford L. Barrett" <brad@...> wrote:
            >
            >
            > Certain malformed user agent strings will cause a segfaut when you
            > use a MangleAgent level of one (1). Has nothing to do with logfile
            > name or where you specify it.
            >

            Thanks for the answer. I tried setting MangleAgent to 0 and also to
            5. I'm still getting the segfault. :(

            Thinking it could be something in the log, I decided to try it with
            just one line in the log:
            xxx.xxx.xxx.xxx - - [01/Apr/2006:00:22:30 -0500] "GET /agent/
            HTTP/1.1" 200 22473

            (I've replaced the real IP address with xxx.xxx.xxx.xxx). Ran the
            following line:

            webalizer-2.01-10/webalizer -d -p -c
            /export/home/stry_cat/w3logs/conf.test/agent/agent.conf -o
            /export/home/stry_cat/w3logs/reports.test/agent -t "agent on " -n
            mydomain.com /export/home/stry_cat/w3logs/data.test/agent/agentAPR06a.log

            And still got the segfault.

            I was about to post the conf file here too, when I noticed that I had
            the HTMLHead command on a line all by itself without anything after
            it. I commented it out and ran my test again. So far it hasn't
            segfaulted in 3 tests. I'm not going to get my hopes up and will post
            again if that doesn't seem to fix the problem.



            > --
            >
            > On Mon, 8 May 2006, Tom Cat wrote:
            >
            > > Well I think I've finally found some more information aboute the seg
            > > faults on RHEL 3&4. The ones that segfault never have a LogFile line
            > > in the configuration file (I always specifiy one on the command line).
            > > The ones that don't segfault have this line, but it gets ignored
            > > since I'm specifying it on the command line.
            > >
            > > Does this help anyone figure out how to fix the problem?
            > >
            > >
            > >
            > >
            > >
            > >
            > >
            > > Webalizer homepage: http://www.webalizer.org
            > >
            > > Yahoo! Groups Links
            > >
            > >
            > >
            > >
            > >
            > >
            > >
            > --
            > Bradford L. Barrett brad@...
            > A free electron in a sea of neutrons DoD#1750 KD4NAW
            >
            > The only thing Micro$oft has done for society, is make people
            > believe that computers are inherently unreliable.
            >
          • Robin Sheat
            ... Have you tried making it segfault while running in strace or gdb? They may give you some hints about what it doing when it dies. -- Robin
            Message 5 of 6 , May 15 11:52 PM
            • 0 Attachment
              On Tuesday 09 May 2006 06:14, Tom Cat wrote:
              > Ok not good folks.  I've just found two cases that invalidated this
              > theory.
              Have you tried making it segfault while running in strace or gdb? They may
              give you some hints about what it doing when it dies.

              --
              Robin <robin@...> JabberID: <eythian@...>

              Hostes alienigeni me abduxerunt. Qui annus est?

              PGP Key 0xA99CEB6D = 5957 6D23 8B16 EFAB FEF8 7175 14D3 6485 A99C EB6D
            • Tom Cat
              ... Just an update on this segfault issue. I ve done the monthly logs for all of our sites using this fix. It does seem like this workaround has solved the
              Message 6 of 6 , Jul 20, 2006
              • 0 Attachment
                --- In webalizer@yahoogroups.com, "Tom Cat" <stry_cat@...> wrote:
                >
                > I was about to post the conf file here too, when I noticed that I had
                > the HTMLHead command on a line all by itself without anything after
                > it. I commented it out and ran my test again. So far it hasn't
                > segfaulted in 3 tests. I'm not going to get my hopes up and will post
                > again if that doesn't seem to fix the problem.
                >

                Just an update on this segfault issue. I've done the monthly logs for
                all of our sites using this fix. It does seem like this workaround
                has solved the problem.

                -Tom
              Your message has been successfully submitted and would be delivered to recipients shortly.