Re: MIME types not delivering

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Thu, Oct 25, 2012 at 04:34:50PM -0400, J.Lance Wilkinson wrote:
> J.Lance Wilkinson wrote:
> >     The file itself is being genrated not via a CGI but some Oracle
> >     process (this IS an Apache configured by Oracle after all), so he's
> >     trying to figure out now just what might be generated by that
> >     procedure, which DOES seem to generate a Content-Type header.  If
> >     we can determine that it gets OMITTED in this case that would explain
> >     the type/plain from the DefaultType, perhaps, even with the DefaultType
> >     directive being removed.   Another possibility is the process itself
> >     has no idea what to assign as a Content-Type, and thus it backpedals
> >     and assigns text/plain because it has nothing better  to assign
> >     (although application/octet would be better, eh?).
> 
> 	I was wondering if there was an apache module or perhaps some obscure
> 	directive that would cause a CGI script or a handler or whatever
> 	the user's Oracle process would be considered to "Tee" out the output
> 	sent back to Apache HTTPD for delivery to the browser and dump that
> 	output into a file.   So we could see what headers if any are
> 	generated by that process.
> 
> 	Thoughts?

If it were using CGI then you could just run the Oracle thing from a
commandline prompt and look at the output there.  You might have to
provide some environment variables to get it to run sensibly:

  http://en.wikipedia.org/wiki/Common_Gateway_Interface

It's kind of hard to debug this without knowing how HTTPD is
communicating with the Oracle thingy.  Could we get a look at the
HTTPD configuration?  Does the Oracle bit have a name?

It ought to be doing one of three things, I think:

o  Some program is run once per request.  It could be a tiny thing that
   just makes contact with the real information provider.  In this
   case I would just make that per-request thing speak CGI; I can't
   imagine a reason to invent some other protocol.  So you could just
   run it and see what it says.

o  HTTPD is a proxy for this other process.  There ought to be
   configuration to reveal that.  If you can figure out the host:port
   for the connection then you could use tcpdump or Wireshark or
   whatever to just snoop the packets.

o  There's (a) persistent connection(s) between HTTPD and the main Oracle
   thing.  In this case there should be an HTTPD module a la mod_dbd
   to set up and run the connection, and there should be configuration
   to lead us to it, or you could at least see it in server_info.  I'd
   guess there would be an AddHandler directive to have it called
   under the right conditions.  You'd probably have to get Oracle to
   tell you how to turn on debug logging.

Come to think of it, if Oracle set it all up then making it work ought
to be one of the reasons you're paying them bales of money.

-- 
Mark H. Wood, Lead System Programmer   mwood@xxxxxxxxx
Asking whether markets are efficient is like asking whether people are smart.

Attachment: pgpm2DpMxwDiP.pgp
Description: PGP signature


[Index of Archives]     [Open SSH Users]     [Linux ACPI]     [Linux Kernel]     [Linux Laptop]     [Kernel Newbies]     [Security]     [Netfilter]     [Bugtraq]     [Squid]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux RAID]     [Samba]     [Video 4 Linux]     [Device Mapper]

  Powered by Linux