We get intermittent "An internal server error occurred. Please try again later" error pages served by pound, and for the life of me I can't figure out what is wrong. Here's a sample of what shows up in the pound log:
Oct 16 04:22:12 pound: (b73dfb90) e500 response error read from XXX / POST / sudoku-lookup.pl HTTP/1.1: Success (0.043 secs) Oct 16 05:34:16 pound: (b69e7b90) e500 response error read from YYY / GET /js/game-common.js?20091008.03 HTTP/1.1: Success (0.000 secs) Oct 16 05:47:22 pound: (b6ebab90) e500 response error read from YYY / GET /js/common.js?20091008.03 HTTP/1.1: Success (0.001 secs) Oct 16 10:31:47 pound: (b7bffb90) e500 response error read from XXX / POST / login.pl HTTP/1.1: Success (0.187 secs) Oct 16 10:41:38 pound: (b79b6b90) e500 response error read from YYY / GET /js/common.js?20091008.03 HTTP/1.1: Success (0.000 secs) Oct 16 12:13:41 pound: (b7628b90) e500 response error read from YYY / GET /css/main_blue.css?20091008.03 HTTP/1.1: Success (0.000 secs) Notice the puzzling "Success" as the error explanation. We have 2 different "services" behind pound -- lighttpd for static files (marked as YYY in the log above), and Apache with mod_perl (XXX in the log) -- getting an equal share of these errors, which makes me think it's probably pound to blame. They are also running on the same physical server, so network errors shouldn't be a problem either. Another puzzling part to this issue is that I cannot find the corresponding requests either in the Apache or lighttpd logs (access or error). Although it could be that I'm simply not finding them because there is no IP address included in pound's error message, but I'm fairly certain they are not there. Can anyone shed some light on what it might be? These errors are rare (maybe one for every 10k-20k requests), but still somewhat worrying. -Mikko -- To unsubscribe send an email with subject unsubscribe to [email protected]. Please contact [email protected] for questions.
