Hi
This might of interest :
http://weblogs.java.net/blog/emcmanus/archive/2006/11/jsr262_and_wsma.html
The idea behind the spec mentioned there is that a JMX WS connector will expose the JMX instrumentation details through a
WS-Management like. There's was a paper on the convergence of various management protocols like WS-management and WSDM and the JMX
WS-Connector spec takes that into consideration...
Cheers, Sergey
----- Original Message -----
From: "Liu, Jervis" <[EMAIL PROTECTED]>
To: <[email protected]>
Sent: Saturday, February 17, 2007 7:43 AM
Subject: RE: Review of patch for CXF-427 [JMX]
Exactly, this is a design choice. We either need to make an abstract layer then build up everything like JMX/SNMP/WSDM etc on top
it, or we just use JMX in the core, then bridge everything else from JMX, for example JMX-SNMP etc.
-----Original Message-----
From: Dan Diephouse [mailto:[EMAIL PROTECTED]
Sent: 2007?2?17? 4:26
To: [email protected]
Subject: Re: Review of patch for CXF-427 [JMX]
FWIW, I have written a JMX->SNMP plugin which is available here:
http://svn.apache.org/repos/asf/geronimo/xbean/sandbox/xbean-jmx/
I have not looked at WSDM yet, but it would be great to know
whether an
abstraction layer is really needed or whether we could just use JMX
throughout. From an SNMP point of view, I don't think its a
requirement that
we have an abstraction layer though. I will be sure to read
the relavent
tuscany discussion though.
Seumas: been meaning to look at your patch, but I haven't had
a chance yet
today. It is open on my desktop though :-)
- Dan
On 2/16/07, Liu, Jervis <[EMAIL PROTECTED]> wrote:
>
> Hi Seumas, sorry I did not follow the previous discussion
of "JMX Changes"
> very closely, I should have raised this concern early. I
believe there is
> still one concern needs to be addressed, which was
mentioned by Willem, but
> has not been paid enough attention to.
>
> Basically, I agree it is possible to simply how CXF
instrumentation works
> by removing EventProcessor/EventListener, but I think we
still need a good
> discussion to decide whether or not we should make the core of CXF
> instrumentation depends on JMX directly. CXF
instrumentation was inherited
> from Celtix, one of core concepts of Celtix instrumentation
design is
> "Instrumentation will have no hard dependency on any
specific management
> protocols, such as JMX", see [1]. To be honest, I am not
sure if this is an
> over design, as Celtix only has JMX supported anyway at the
end of day, and
> I do not see CXF has any immediate requirement to support management
> protocols other than JMX, such as SNMP and WSDM. But this
kind of capability
> definitely appears interesting to me, and it is worth a
good consideration
> during the design, though the "common management model"
approach proposed by
> Celtix may not be the only way to achieve the goal. Back to
couple months
> ago, I was involved in a discussion in Tuscany on the topic of SCA
> management where they have a requirement to support both
JMX and WSDM for
> Tuscany, see [2] [3] [4].
>
>
> [1].
https://wiki.objectweb.org/celtix/Wiki.jsp?page=ManamgentDevPlan
>
> [2].
http://www.mail-archive.com/[email protected]/msg03625.html
>
> [3].
http://www.mail-archive.com/[email protected]/msg04025.html
>
> [4].
http://www.mail-archive.com/[email protected]/msg04025.html
>
> Cheers,
> Jervis
>
> > -----Original Message-----
> > From: Soltysik, Seumas [mailto:[EMAIL PROTECTED]
> > Sent: 2007?2?16? 7:32
> > To: [email protected]
> > Subject: Review of patch for CXF-427 [JMX]
> >
> >
> > I just uploaded a patch file for Jira CXF-427 which involves
> > a refactoring of the current JMX infrastructure. Could
> > someone take a look at this patch and apply it if deemed
sufficient.
> > Thanks,
> > Seumas
> >
>
--
Dan Diephouse
Envoi Solutions
http://envoisolutions.com | http://netzooid.com/blog
