On 11/16/06, Dan Fandrich <dan_at_coneharvesters.com> wrote:
> On Thu, Nov 16, 2006 at 04:18:44PM -0800, Demetri Mouratis wrote:
> > Thanks for the reply. Here's some further detail based on your suggestions.
> >
> > [root_at_dropbox4 ~]# /usr/bin/curl -s -v -x pakiti.scc.rnmd.net:80 -F
> > host="dropbox4.lnc.rnmd.net" -F kernel="2.6.9-42.ELsmp" -F
> > admin="dmourati_at_rnmd.net" -F os="CentOS release 4.4 (Final)" -F
> > version="5" -F updates="\<UPDATES\>\<\/UPDATES\>"
> > http://pakiti.scc.rnmd.net/feed/
> > * About to connect() to pakiti.scc.rnmd.net port 80
> > * Trying 207.0.21.137... * connected
> > * Connected to pakiti.scc.rnmd.net (207.0.21.137) port 80
> > >POST http://pakiti.scc.rnmd.net/feed/ HTTP/1.1
> > User-Agent: curl/7.12.1 (i686-redhat-linux-gnu) libcurl/7.12.1
> > OpenSSL/0.9.7a zlib/1.2.1.2 libidn/0.5.6
> > Host: pakiti.scc.rnmd.net
> > Pragma: no-cache
> > Accept: */*
> > Content-Length: 714
> > Expect: 100-continue
> > Content-Type: multipart/form-data;
> > boundary=----------------------------5f0233e0ae1e
> >
> > < HTTP/1.1 100 Continue
> > < HTTP/1.1 200 OK
>
> This is a completely bogus reply. The server or the proxy is returning two
> HTTP return codes, when it should only send one.
>
> > < Date: Fri, 17 Nov 2006 00:15:45 GMT
> > < Server: Apache/2.0.52 (CentOS)
> > < X-Powered-By: PHP/4.3.9
> > < Content-Length: 16
> > < Connection: close
> > < Content-Type: text/html; charset=UTF-8
> > * Closing connection #0
> > Connection error[root_at_dropbox4 ~]# which curl
>
> The "Connection error" you are seeing is actually text returned by the server
> (or your proxy), and is not a curl error message.
>
> It looks like a bug in your proxy server or HTTP server that's causing this
> failure. You might be able to get around it by using the -0 option to force
> use of HTTP 1.0 and bypass the 100-continue mode that is likely the buggy
> part of the server.
>
Dan

Thanks again. The -0 option gave the same results. The server is
httpd 2.0.52 from CentOS with an installation of the pakiti server. I
gather from your comments that I should take this up with the pakiti
author at this time. I'm fairly confident at httpd itself is not
blame here.

More details to follow if and when I figure this out.

Your prompt help is greatly appreciated.

-D
Received on 2006-11-17