https://bz.apache.org/bugzilla/show_bug.cgi?id=59220

--- Comment #3 from Scott Nicklous <nisco...@googlemail.com> ---
Hi Violeta and Remy,

thank you very much for having a look at this so quickly (and thank you, Remy
for fixing 59213 so promptly!).  The example servlets I provided were for the
purpose of reproducing the problem rather than for illustrating the use case. I
would like to explain my use case more clearly.

I'm working on the Apache Pluto project, which is the reference implementation
for JSR 286 Portlet Specification 2.0, and which will be the reference
implementation for JSR 362 Portlet Specification 3.0. Pluto is a minimal portal
server that implements the required API for portlet applications. 

For version 3.0, we want to provide async support for portlet applications. The
portlet specification can make recommendations about how portlet applications
should do error handling, but can't really guarantee that the portlet apps will
follow the recommendations.

The Pluto server has to allocate resources to support the portlet applications
during async processing and needs a reliable way to release those resources
even if a timeout or error condition occurs. It seems to me that the natural
way to do that would be for the Pluto server to register an AsyncListener on
behalf of the portlet application in order to release the resources when
onComplete() is called.

However, this bug along with 59219 means that in these edge cases, Pluto will
not be able to release the resources, which could potentially result in a
memory leak. 

That said, I understand that these really are edge cases, and I can't really
say how high the priority should be in the grand scheme of things. But it would
be very nice if they could be fixed at some point as time & priority allows.
:-)

thanks again,
Scott

-- 
You are receiving this mail because:
You are the assignee for the bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@tomcat.apache.org
For additional commands, e-mail: dev-h...@tomcat.apache.org

Reply via email to