Re: Make FastCGI reliable
Janis Volbergs <[email protected]>
| Newsgroups | gmane.comp.web.fastcgi.devel |
|---|---|
| Message-ID | <[email protected]> |
Pretty much seems like a segfault. Frank de Bot wrote: >All of sudden the PHP FastCGI failed without any apperent reason: >First I get this in the apache errorlog: > >terminated by calling exit with status '0' > >following by a very long row of > >has failed to remain running for 30 seconds given 3 attempts, its >restart interval has been backed off to 600 seconds > >meanwhile my site was just down :-/ > >A core file was again created > >(gdb) bt >#0 0x48c7d95b in memcpy () from /usr/lib/libc_r.so.4 >#1 0x60796863 in ?? () >#2 0x46233d72 in ?? () >Cannot access memory at address 0x6f6c6f63. > > >At the moment I still can't have this running reliable. > >Any one other hints to prevent this weird situation? > > >Regards, > >Frank de Bot > > > > >Frank de Bot wrote: > > >>Hi, >> >>At the moment I'm running Apache 2.0.58 with mod_fastcgi 2.4.2 . I run >>php 5.1.4 as FastCGI application. The OS is FreeBSD 4.11-stable >>But I can't manage to let it run perfectly. >>I've ran it in 2 ways so far: >>- PHP binary directly as FastCGI >>- PHP Wrapped with a shell script as FastCGI (php launches a few child >>of it own) >> >>The first method will result on a given moment in a crash "terminated >>due to uncaught signal '10' ((null))" and fails to restart until apache >>itself is restarted. This is seen after the php FastCGI has reached the >>set up dynamicMaxClassProcs. backtrace with gdb results in: >> >>#0 0x48c7d95b in memcpy () from /usr/lib/libc_r.so.4 >>#1 0x0 in ?? () >> >>The second method will cause the first launced php script to spawn >>childs without reason and out of control. But request are being >>continued to be served. >>Occasionaly a request will be failed to served, but it does not seem >>related to the uncontrolled spawning of php child processes. >> >>PHP developers had rewritten the fastcgi in php as of 5.1.3. All >>metioned cases above are equally present in versions before and after >>5.1.3. I think a possible cause can be found in fastcgi itself. >> >>How can I make FastCGI more reliable so it handles ALL requests >>properly? Visitors of my website are pretty fed up with 500 errors... >> >> >>Thanks in advanced, >> >> >>Frank de Bot >> >> >> >>___________________________________ >>fastcgi-developers mailing list >>http://fastcgi.com/fastcgi-developers/ >> >> >> > >___________________________________ >fastcgi-developers mailing list >http://fastcgi.com/fastcgi-developers/ > > ___________________________________ fastcgi-developers mailing list http://fastcgi.com/fastcgi-developers/