Re: [cometd-users] Strange behaviour from the cometd long pooling

[Simone Bordet]

Hi,

On Sat, Jan 19, 2013 at 4:21 PM, Boris Horvat <horvat....@gmail.com> wrote:
> Hi all,
>
> I am writing an application in tapestry and I am using tapestry-cometd for
> it. The first time I have integrated this I run into a problem where I could
> not run the application under Tomcat 7 64bit and I needed to revert to 32bit
> version (for more information please look atstackoverflow).
> After that I believe that my application reverted to long pooling as I can
> now see messages in my logs like the one given bellow
>
> 127.0.0.1 - - [15/Jan/2013:23:00:53 +0000] "POST /cometd/connect HTTP/1.1"
> 200 147
>
> Also when I try to go to that address it throws an error 400 Unknown Bayeux
> Transport Exception.
>
> I am not really sure what is the problem, and why the application didnt work
> on 64 version of the Tomcat7 so if anyone can point me in the right
> direction I would appriciated it.
>
> Also for more information about the exception please look at this

Looks like what you're using is Atmosphere, not CometD.

There are no problems that I know in running CometD with Jetty (32 or 64 bit).

You have several (too many ?) layers that seems to cause problems for you.
Perhaps a simpler approach will give you better results ?

Simon
--
http://cometd.org
http://webtide.com
Developer advice, training, services and support
from the Jetty & CometD experts.
----
Finally, no matter how good the architecture and design are,
to deliver bug-free software with optimal performance and reliability,
the implementation technique must be flawless. Victoria Livschitz

[Boris Horvat]

Strange they said the same thing only revers :) (ie that I am using cometd and not atmosphere). 

I am using tomcat 7 would this make a difference? 

Also did you ever encounter anything that could explain the tomcat 7 fail to startup (I think that there was a similar thread about startup fail but it didnt help me)

[Simone Bordet]

Hi,

On Sat, Jan 19, 2013 at 10:40 PM, Boris Horvat <horvat....@gmail.com> wrote:
> Strange they said the same thing only revers :) (ie that I am using cometd
> and not atmosphere).

Who says that ?

Look yourself at the stack trace you posted on stackoverflow:

Caused by: javax.servlet.ServletException:
java.lang.IllegalArgumentException: Comet support class
org.atmosphere.container.JettyAsyncSupportWithWebSocket has bad
signature.
at org.atmosphere.cpr.AtmosphereFramework.init(AtmosphereFramework.java:549)
at org.atmosphere.cpr.AtmosphereServlet.init(AtmosphereServlet.java:191)
at org.atmosphere.cometd.CometdServlet.init(CometdServlet.java:50)

As you can see the packages refer to org.atmosphere, not org.cometd.
So you're using Atmosphere, not CometD, no matter what anyone is saying.
The fact that Atmosphere calls some of their classes "Cometd" it's
strange, though. No idea why.

> I am using tomcat 7 would this make a difference?

It is known to be buggy, depending on the version (but more or less we
have report for almost any version).

> Also did you ever encounter anything that could explain the tomcat 7 fail to
> startup (I think that there was a similar thread about startup fail but it
> didnt help me)

"Fail to startup" is too little information, sorry.