the possibility to manage an access log compatible with apache combined
log format. reopening files is also handled. Also, this change fix
issues #237 and #190.
For the sake of simplicity logging file have been removed
--logfile option has become --error-logfile, by default "-"
Following options have been added:
--access-logfile, the access log file path. By default None. If set
access logs will bre written to defined output
eg:
$ gunicorn -w3 --access-logfile=- test:app
2011-08-21 19:36:52 [84572] [INFO] Starting gunicorn 0.13.0
2011-08-21 19:36:52 [84572] [INFO] Listening at: http://127.0.0.1:8000
(84572)
2011-08-21 19:36:52 [84572] [INFO] Using worker: sync
2011-08-21 19:36:52 [84573] [INFO] Booting worker with pid: 84573
2011-08-21 19:36:52 [84574] [INFO] Booting worker with pid: 84574
2011-08-21 19:36:52 [84575] [INFO] Booting worker with pid: 84575
127.0.0.1 - - [21/Aug/2011:19:37:04] "GET / HTTP/1.1" 200 14 "-"
"curl/7.21.4 (universal-apple-darwin11.0) libcurl/7.21.4 OpenSSL/0.9.8r
zlib/1.2.5"
This can cause odd behavior with browsers not being able to parse the response
correctly (not sure if this is just a property of empty chunked responses or
something else)
`urlparse.urlparse` splits the url at semicolons and calls it params. Since
the WSGI doesn't have any special treatment for params, they should be
preserved and passed to the wsgi app. Using `urlparse.urlsplit` to
avoid splitting the params.
StopIteration is simply propogated, so there's no reason to catch it.
The catch-all exception is also caught and logged in handle(), but
broken pipe errors are ignored. The redundant catch here was generating
excess logging by calling handle_error for a broken pipe which fails to
send the error because the client has already disconnected.
Instead of hardcoding X-Forwarded-Protocol and X-Forwarded-SSL, make the
header and value configurable, with no default that would enable a
client to spoof secure requests if the reverse proxy is not configured
to strip the header used.
aren't correctly reloaded. First attempt to remove it from sys.modules
introduced more problem than it solved.
This new attempt, just get the module, reload it and populate currently
loaded settings with new values.
Its possible that when iterating Arbiter.WORKERS in manage_workers we
get interupted to handle a SIGCHLD which will pop the child PID from
the dict which results in a "dict changed size while iterating
error. Reported on IRC. Simple fix is to just copy the dict into a
list that we iterate.
Move "import gevent.wsgi" into a separate module, which is not
imported if pywsgi worker is requested. (The gevent.wsgi currently
not provided by gevent trunk but pywsgi works fine.)
