pipe delivery process timed out

ifbtech

• November 08, 2013 17:30

I'm getting this error in exim_mainlog for a couple pipes: 2013-11-08 09:23:27 1Vemy7-0000JQ-EJ <= sendinguser@domain.org H=h58.185.17.198.static.ip.windstream.net (HP) [198.17.185.58]:60255 P=esmtpa A=dovecot_login:sendinguser@domain.org S=6297 i d=03f301cedc8e$17444560$45ccd020$@org T="Good Morning" for goodmorning@domain.org 2013-11-08 09:23:27 cwd=/var/spool/exim 3 args: /usr/sbin/exim -Mc 1Vemy7-0000JQ-EJ 2013-11-08 10:23:27 1Vemy7-0000JQ-EJ ** |/home/domain/public_html/mt/pipe-goodmorning.php (goodmorning@domain.org) R=virtual_aliases_nostar T=jailed_virtual_address_pipe: pipe delivery process timed out 2013-11-08 10:23:27 cwd=/var/spool/exim 7 args: /usr/sbin/exim -t -oem -oi -f <> -E1Vemy7-0000JQ-EJ 2013-11-08 10:23:27 1VenuB-0004NI-NT <= <> R=1Vemy7-0000JQ-EJ U=mailnull P=local S=7188 T="Mail delivery failed: returning message to sender" for sendinguser@domain.org 2013-11-08 10:23:27 cwd=/var/spool/exim 3 args: /usr/sbin/exim -Mc 1VenuB-0004NI-NT 2013-11-08 10:23:28 1Vemy7-0000JQ-EJ Completed 2013-11-08 10:23:28 1VenuB-0004NI-NT => sendinguser R=virtual_user T=virtual_userdelivery 2013-11-08 10:23:28 1VenuB-0004NI-NT Completed My PHP pipe file starts with: #!/usr/bin/php -q The php process runs for an hour and then is killed and the bounce message sent to the sender. The actual processing is complete in less than a minute. I had it save to a log file and that file shows it's done with a timestamp of Nov 8 09:23. These pipes have worked fine for years and then I moved the account to a new server with most everything the same. PHP is newer, I'm using FastCGI now where I was using mod_php before but I don't think that should affect this at all. I've tried adding set_time_limit(600) to get the PHP process to just close after 10 minutes but it doesn't work. I've tried adding an exit; to make sure it dies but no luck. Thanks in advance for the help!

• 

  cPanelMichael

  • November 12, 2013 17:39

  Hello :) I see you have opened a support ticket for this issue. I will update this thread with the outcome once the ticket is closed. Thank you.

  0
• 

  ifbtech

  • November 14, 2013 16:15

  Thanks, here's the solution: /usr/bin/php is intended for use with the web server, and /usr/local/bin/php is the CLI version. I suspect that your previous server must have been compiled without FastCGI support, meaning that /usr/bin/php on that server would have been configured with a different SAPI than the current server. A FastCGI PHP binary will process its first request, and then sit open, waiting for additional requests, until it's killed. I changed the shebang to /usr/local/bin/php and now it's all good!

  0
• 

  cPanelMichael

  • November 14, 2013 17:15

  I am happy to hear the issue is now resolved. Thank you for updating this thread with the solution.

  0

Please sign in to leave a comment.