Mailing-List: contact tomcat-dev-help@jakarta.apache.org; run by ezmlm
Precedence: bulk
Reply-To: tomcat-dev@jakarta.apache.org
Message-ID: <3B55E0CF.F47B25DA@openwave.com>
Date: Wed, 18 Jul 2001 12:17:35 -0700
From: "Forrest R. Girouard" <Forrest.Girouard@openwave.com>
MIME-Version: 1.0
To: tomcat-dev@jakarta.apache.org
Subject: Re: Tomcat & Frames
References: <Pine.BSF.4.21.0107181007520.39118-100000@localhost>
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit


I could be wrong but I thought that if the server supports HTTP 1.1
and the client does not receive a 'connection: close' with the response
that it will re-use the connection.  This means that some of the multiple
simultaneous requests might be handled by the same connection and if the
server has a bug in it's handling of persistent connections one might
get an unexpected "connection reset by peer".   So, if all the other 
possibilities can be eliminated I would investigate the handling of
persistent connections.

Cheers,
	Forrest


"Craig R. McClanahan" wrote:
> 
> Servers don't know anything about frames.  The only relevance it has in
> this context is the fact that your browser will typically do multiple
> requests (one per frame) simultaneously, up to some maximum -- typically
> 4.  This is the same reason you see multiple images being downloaded
> simultaneously (these are also done on separate requests).  In most
> browsers, the number of simultaneous requests is configurable somewhere.
> 
> The server is reporting "connection reset by peer" in the middle of a
> request.  That means one of the following things:
> * The browser started a request and the user pressed STOP.
> * The browser started an request in one frame and
>   JavaScript on another frame told it to go somewhere else (which
>   has the same effect as a user pressing STOP).
> * You have a meta refresh reload someplace that is firing so fast that
>   the previous request hasn't had a chance to complete yet.  The five
>   second periodicity makes me pretty suspicious of this.
> * You have some network component in between (a router or proxy) that
>   is misbehaving.
> * You have some completely separate connection attempts going on that
>   are disobeying the HTTP protocol (such as doing a telnet connection
>   to port 8080 and then disconnecting).
> 
> Craig
> 
> On Wed, 18 Jul 2001, Cory Powers wrote:
> 
> > That is what I would have thought to but, as I noted in the bug report, I
> > can reproduce the error in IE and Netscape on multiple platforms(different
> > versions of Windows and Linux). Please take a look at this, I think there is
> > a real issue here. I'm trying to investigate deeper but this is my first
> > expirence with tomcat so I need figure out the code before I could really
> > make any progress. Any pointers on where to look would be greatly
> > appreciated.
> >
> > BTW I have posted this to the user list and recieved reports of other users
> > having the same issue but, the only resolution was to restart the server and
> > this did not work for me, strange...
> >
> > Thanks,
> > Cory
> > cpowers@appriss.com
> >
> > -----Original Message-----
> > From: Craig R. McClanahan [mailto:craigmcc@apache.org]
> > Sent: Wednesday, July 18, 2001 11:59 AM
> > To: 'tomcat-dev@jakarta.apache.org'
> > Subject: Re: Tomcat & Frames
> >
> >
> > First, because servlet containers (and web servers, for that matter) don't
> > have a clue about frames, it's 99.9% sure to be either an application
> > error, a browser bug, or else a harmless warning that the browser
> > stopped downloading one frame because some JavaScript in the page told it
> > to do so.
> >
> > Second, these sorts of issues are better discussed on the TOMCAT-USER
> > mailing list.
> >
> > Craig McClanahan
> >
> > On Mon, 16 Jul 2001, Cory Powers wrote:
> >
> > > I am having issues using Tomcat with a frameset page that contains 4
> > frames,
> > > I do not have any problems with the same page when I reduce it to 3
> > frames.
> > > I have created a Bug in the Bugs Database(Bug #2605) and gotten no
> > response.
> > > Can someone please assist me in getting this solved. The Bug entry has a
> > lot
> > > more detail.
> > >
> > > Thanks,
> > > Cory
> > > cpowers@appriss.com
> > >
> >

-- 
Forrest Girouard @ Openwave Systems Inc.
phone: +1-650-480-4184
mailto:Forrest.Girouard@openwave.com
http://www.openwave.com