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 >
