httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jeff Trawick" <traw...@gmail.com>
Subject Re: Palm Treo access to OWA via Apache 2.2.x Proxy
Date Thu, 22 May 2008 22:28:43 GMT
On Thu, May 22, 2008 at 4:59 PM, Ray Van Dolson <rvandolson@esri.com> wrote:
> On Thu, May 22, 2008 at 01:52:11PM -0700, Jim Jagielski wrote:
>>
>> On May 21, 2008, at 9:05 PM, Ray Van Dolson wrote:
>> > As I said, everything works fine this way.  With mod_proxy in the mix
>> > however:
>> >
>> >   1 TREO   Sends OPTIONS request to Apache (Connection: Keep-Alive)
>> >   2 Apache Sends OPTIONS request to IIS (Connection: Close)
>> >   3 IIS    Replies with 200 OK (Connection: Close)
>> >   4 Apache Sends 200 OK back to Treo (Connection: Close)
>> >   5 Apache Sends SSLv3 Close Notify to Treo
>> >   6 Apache Sends FIN,ACK to Treo
>> >   7 TREO   Sends POST /Microsoft-Server-ActiveSync... (Connection:
>> > Keep-Alive)
>> >   8 Apache Sends ACK to Treo
>> >   9 Apache Sends Dup ACK to Treo
>> >  10 Apache Sends Dup ACK to Treo
>> >
>> > Timeout...
>> >
>> > I attempted to backport the patches mentioned here:
>> >
>> >  http://svn.apache.org/viewvc?view=rev&revision=657440
>> >
>> > In the hopes it might address this, but I'm still seeing the
>> > connection
>> > prematurely closed around #5, #6 above.  I *believe* this is the cause
>> > of my issue but am not positive.  The Treo continues to try and use an
>> > existing connection that no longer exists, and Apache keeps sending
>> > Connection: close headers.
>> >
>>
>> I don't see how this is a bug. Why does the TREO not
>> close the socket and reopen another? How is it "premature"?
>> The server can decide to say "That's enough" at any time and
>> the client must honor that. Instead, it looks like the
>> TREO is seeing 'Connection: close' and ignoring it and
>> considering the connection persistent.
>
> Understood; I'm happy to blame the Treo for perhaps poorly implementing
> HTTP, but from the user perspective it works directly to IIS, but not
> via the Proxy.
>
> My most feasible option is to try and see if this is something that can
> be addressed within Apache.  An option to allow a persistent connection
> in the scenario above might help.

Try to get the Treo into different connection-handling logic
altogether and see if that helps.  Maybe force-response-1.0 will do
it. (http://httpd.apache.org/docs/2.2/env.html#special)

> I promise to go and read through the RFC's, but if the Treo is
> requesting a Keep-Alive connection, shouldn't Apache try its best not
> to close the connection as quickly as it is?

no ;)

Mime
View raw message