modwsgi

[Prev] Thread [Next]  |  [Prev] Date [Next]

[modwsgi] Re: Sometimes my Django site just does not respond at first Graham Dumpleton Thu Jun 18 23:00:38 2009

2009/6/19 Robert Coup <[EMAIL PROTECTED]>:
> On Fri, Jun 19, 2009 at 3:49 PM, Robert Coup <[EMAIL PROTECTED]>
> wrote:
>>
>> I'm looking into this error as well. Trying to correlate it with
>> requests/uploads/reloads/etc to figure out what's going
>> on. More info as i find it...
>
> Here's a case:
> Main Apache Error Log:
> [Fri Jun 19 05:23:06 2009] [error] [client 123.123.123.123] mod_wsgi
> (pid=29519): Exception occurred processing WSGI script
> '/etc/myapp/apache/myapp.wsgi'.
> [Fri Jun 19 05:23:06 2009] [error] [client 123.123.123.123] IOError: client
> connection closed
>
> VHost Log:
> [Fri Jun 19 05:23:06 2009] [error] [client 123.123.123.123] (104)Connection
> reset by peer: mod_wsgi (pid=29580): Unable to get bucket brigade for
> request., referer: http://example.com/my/webupload/
> In the access log:
> 123.123.123.123 - - [19/Jun/2009:05:22:06 +0000] "POST
> /my/webupload/?token=1245388915969 HTTP/1.1" 500 675
> "http://example.com/my/webupload/" "Mozilla/5.0 (Windows; U; Windows NT 5.1;
> en-GB; rv:1.9.0.6) Gecko/2009011913 Firefox/3.0.6 (.NET CLR 3.5.30729)"
> Which occurred in the log at ~05:23:06 (60s later)
> Config stuff:
> - Apache Timeout is set to 300 in this case
> - the wsgi daemons are single-threaded
> - the deadlock-timeout is set to 90 (but increasing to 300 didn't help)
> - Apache LogLevel for the VHost was set to info
> this all happened just after a clean apache restart on that server, and the
> file being uploaded was ~10MB.

But was still random, or was it reproducible to some degree?

> Other people seem to be able to complete uploads happily... so i'm a bit
> puzzled.

For this upload case, have you use curl to do the upload, or some
other way which has a means to monitor the progress of the upload so
it is possible to see whether upload progress stops some time before
the error is raised. Obviously if it isn't readily reproducible, that
is going to be hard to do.

Anyway, your post has reminded that the unable to get bucket brigade
error is just a lower level version of my higher level client closed
connection error. Ie., duplicated error logging effectively.

Graham

--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups 
"modwsgi" group.
To post to this group, send email to [EMAIL PROTECTED]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/modwsgi?hl=en
-~----------~----~----~----~------~----~------~--~---