Oleg, Roland,
As suggested by Roland, Connection: Close fixed the problem. It had
the same effect on the wire communication as if a new HttpClient were
used for each request. So, I'm not sure what the issue was exactly,
but keep-alive isn't that important for me right now, so I've got a
solution. I'll talk to the Jetty guys to see if we can figure the
root cause out later.
Thanks for the tips!
Patrick Lightbody
Autoriginate, Inc.
503-488-5402
http://www.autoriginate.com
[EMAIL PROTECTED]
"Intelligent testing made convenient"
On Aug 8, 2006, at 3:28 PM, Oleg Kalnichevski wrote:
On Tue, 2006-08-08 at 10:23 -0700, Patrick Lightbody wrote:
More info:
netstat on the client that is making the HttpClient call:
tcp 0 0 ::ffff:192.168.167.1:48990 ::ffff:
192.168.167.133:8081 ESTABLISHED
netstat on the server that is accepting the call:
TCP node-x:8081 192.168.167.1:48990 ESTABLISHED
Complete stack dump of the JVM that is accepting the call (I see no
active socket connection here):
Patrick,
This goes to support my suspicion that the problem is on the server
side, as the server is not blocked expecting data from the client but
rather appears to be blocked/busy doing something else.
...
And finally, XFire reuses HttpClient objects. When I forced it not
to, everything worked fine. What does that mean?
Feel free to post wire logs of both sessions (one with connection
reuse
and one without) and I'll try to see if I can spot anything unusual
Oleg
Patrick
Patrick Lightbody
Autoriginate, Inc.
503-488-5402
http://www.autoriginate.com
[EMAIL PROTECTED]
"Intelligent testing made convenient"
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: httpclient-user-
[EMAIL PROTECTED]
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
