"Seth Milder" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Hi list,
>
> I am seeing this in the logs when posting to a particular URL:
>
> [Mon Apr 03 14:52:06 2006] [13089:15648] [debug]
> ajp_get_endpoint::jk_ajp_common.c (2131): acquired connection cache slot=0
> [Mon Apr 03 14:52:06 2006] [13089:15648] [error]
> ajp_marshal_into_msgb::jk_ajp_common.c (490): failed appending the SSL
> certificates
> [Mon Apr 03 14:52:06 2006] [13089:15648] [info]
> ajp_service::jk_ajp_common.c (1662): Creating AJP message failed,
> without recovery
> [Mon Apr 03 14:52:06 2006] [13089:15648] [debug]
> ajp_done::jk_ajp_common.c (2074): recycling connection cache slot=0 for
> worker ajp13
> [Mon Apr 03 14:52:06 2006] [13089:15648] [info] jk_handler::mod_jk.c
> (1978): Aborting connection for worker=ajp13
>
>
> It is then I receive the apache error about the "Request Entity Too
> Large." Interestingly enough, this only occurs with Firefox and not
> MSIE. Is there a setting for this somewhere?
>
Probably not. The problem is that AJP/1.3 only allows a maximum of 8Kb to
be sent in a given packet. For the Request packet this has to include all
of the Request headers, as well as the SSL variables. I'm guessing (since
you say you are doing a POST), that the referer URL is very long, and that
FireFox is adding just a bit more than MSIE to push it over the edge.
If it's possible for you to hack your copy of the mod_jk code, then putting
something like:
jk_dump_buff(l, __FILE__, __LINE__, "ajp_marshal_into_msgb",
JK_LOG_LEVEL_TRACE, "ssl_cert", msg);
at line 490 of jk_ajp_common.c, and configuring your log-level to trace will
give you the info to see what is wrong. And feel free to post the results
to the list: Myself and at least a couple of other people can parse what
the data in an AJP/1.3 message is.
>
> Thanks!
>
> Seth
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
