503 Errors when accessing folder via alias

[neekster]

We are having an issue accessing aliases, with attempts giving a 503 error but really struggling to pin this one down. Here's what we are doing.

We are running DirectAdmin with LiteSpeed 4.2.10, as it's running on top of DirectAdmin, litespeed is using Apache's config.

There's an apache alias for /phpmyadmin setup by default with DA that points to /phpMyAdmin-4.1.6-all-languages

Now whenever we access /phpmyadmin we get a 503 error, but accessing it via /phpMyAdmin-4.1.6-all-languages works fine.

This is what we get in the error logs

2014-05-14 23:26:38.384 [INFO] [lsphp5] PID: 23537, add child process pid: 23583, procinfo: 0x2359d40
2014-05-14 23:26:38.408 [INFO] Remove pid: 23583, exitcode: 255
2014-05-14 23:26:38.408 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] connection to [/tmp/lshttpd/lsphp5.sock] on request #0, confirmed, 0, associated process: -1, running: 0, error: Connection reset by peer!
2014-05-14 23:26:38.416 [INFO] [lsphp5] PID: 23537, add child process pid: 23584, procinfo: 0x2359d40
2014-05-14 23:26:38.441 [INFO] Remove pid: 23584, exitcode: 255
2014-05-14 23:26:38.441 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] connection to [/tmp/lshttpd/lsphp5.sock] on request #0, confirmed, 0, associated process: -1, running: 0, error: Connection reset by peer!
2014-05-14 23:26:38.449 [INFO] [lsphp5] PID: 23537, add child process pid: 23585, procinfo: 0x2359d40
2014-05-14 23:26:38.474 [INFO] Remove pid: 23585, exitcode: 255
2014-05-14 23:26:38.474 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] connection to [/tmp/lshttpd/lsphp5.sock] on request #0, confirmed, 0, associated process: -1, running: 0, error: Connection reset by peer!
2014-05-14 23:26:38.474 [NOTICE] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] Max retries has been reached, 503!
2014-05-14 23:26:38.474 [NOTICE] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] oops! 503 Service Unavailable
2014-05-14 23:26:38.474 [NOTICE] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] Content len: 0, Request line: 'GET /phpMyAdmin/ HTTP/1.1'
2014-05-14 23:26:38.474 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] Cookie len: 167, session=6TaLpwnuJzMJ31IYzLVZYu4n2Z16fG3jhc2UWoOpli9oiPKhkpEJhI4jr5nULpB0; LSWSWEBUI=b80b634253fe67e564d767e1a90cd835; lsws_uid=WJwuKElWkMU%3D; lsws_pass=ZeWaG7asUco%3D
2014-05-14 23:26:38.474 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] abort request..., code: 4
2014-05-14 23:26:38.474 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] abort request..., code: 4
2014-05-14 23:26:38.474 [INFO] [209.126.225.188:48582-0#APVH_110.232.112.176:443_localhost] File not found [/var/www/html/503.shtml]

But the above is not making all that much sense to me. I've compared LiteSpeeds configuration to one of our other servers that works fine, and I can't find anything significantly different and everything is running in a reasonably stock configuration.

Anybody got any ideas? I'm lost.

Thanks

 

[mistwang]

That means that lsphp5 process crashed when you access /phpmyadmin, you can check stderr.log or PHP error_log if you can locate that.

When you access /phpMyAdmin-4.1.6-all-languages directly, different set of lsphp5 process was used. Maybe one lsphp5 process group is in suEXEC mode, the other one is in non-suEXEC mode.