Squid Support (Henrik Nordstrom) wrote:
> Sorry, I am not sure I understand what you are doing here or exacly
> when you are getting problems. Are you saying that you cannot succeed
> in getthing these replies at all back via Squid, or that it does not
> work to keep the connection persistent for a second request?
- It works fine to send many requests (where each request sends its
little response) (with Squid)
- It works fine to send a single request and return many responses.
(with Squid)
- It does NOT work to send a request which sends many responses and
this multiple times (with Squid)
Without squid all cases run fine.
>
> What I can tell is that the reply headers you have shown can't keep a
> HTTP/1.0+ persistent connection, and will be closed when the reply is
> finished. HTTP/1.0+ persistent connections require a Content-length
> reply header. (HTTP/1.1 has a mode where you can keep connections
> persistent without the need of Content-length, but Squid is not yet
> HTTP/1.1 and neither is your reply)
What is wrong with the reply?
OK, what do i have to do?
Does this help?
Content-Length: 100000000000
I can send two test java files which show the problem
(if you are willing to look at it :-)
I have just download httptunnel-3.0.5, there
it works - how can i dump the squid messages
to see how they have solved it?
thanks for your help,
Marcel
>
> Regards
> Henrik Nordström
> Squid Developer
>
>
>
> On Wednesday 20 February 2002 10:53, Marcel Ruff wrote:
>
>>Hi,
>>
>>
>>My problem is as follows:
>>
>>
>>client --- squid ----- server
>>
>>
>>I open a scoket connection from client to proxy, with a
>>
>> POST http://server ...
>>
>>this works fine.
>>
>>1) Now i can send through this socket many request (works fine).
>>
>>2) Then i tried to send a request and push many data packages
>>back to the client (works fine).
>>
>>Now i try to do 1) and 2) simultaneously, but the proxy
>>kicks me off (the code works fine without proxy).
>>To be more precise, i believe the socket connection is
>>still there, but the data written to the socket is not
>>reaching the counterpart.
>>
>>This is not timeout related as the test runs through in some
>>milliseconds.
>>
>>
>>
>>Now my question:
>>
>>Is there any trick that proxies allow what i want
>>to do (setting some strange http header etc.)?
>>
>>Can i dump more info with squid to find out
>>why the pushed message suddenly does not reach the
>>destination?
>>
>>
>>thanks,
>>
>>Marcel
>>
>>
>>Some code:
>>
>>------------ The request header ----------------------
>>POST / HTTP/1.0
>>From: xx@yy.com
>>User-Agent: HTTP xmlBlaster/1.0
>>Pragma: no-cache
>>Content-Type: text/plain
>>Content-Length: 16
>>Via: 1.1 develop.ruff.de:3128 (Squid/2.4.STABLE2)
>>X-Forwarded-For: 192.168.1.2
>>Host: 192.168.1.2:8080
>>Cache-Control: no-cache, max-age=259200
>>Connection: keep-alive
>>
>>-------------------------------------------
>>
>>
>>------------ The sent callback header ----------
>>HTTP/1.1 200 OK
>>Server: HTTP xmlBlaster server/1.0
>>Cache-Control: no-cache
>>Pragma: no-cache
>>Connection: Keep-Alive
>>Keep-Alive: 30000000
>>Content-Type: multipart/x-mixed-replace;boundary=End
>>
>>-------------------------------------------
>>
>>
>>----------- The received callback header -----
>>HTTP/1.1 200 OK
>>Server: HTTP xmlBlaster server/1.0
>>Cache-Control: no-cache
>>Pragma: no-cache
>>Connection: Keep-Alive
>>Keep-Alive: 30000000
>>Content-Type: multipart/x-mixed-replace;boundary=End
>>
>>-------------------------------------------
>>
>
-- Marcel Ruff mailto:ruff@swand.lake.de http://www.lake.de/home/lake/swand/ http://www.xmlBlaster.orgReceived on Wed Feb 20 2002 - 06:10:07 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 17:06:27 MST