I solute to Anne on her provision! - Michael
----- Original Message ---- From: Anne Thomas Manes <[EMAIL PROTECTED]> To: [email protected] Sent: Friday, October 10, 2008 1:31:15 PM Subject: Re: [service-orientated-architecture] ITIL & SOA Governanc I also predicted circa 2006 that UDDI and CMDB would converge, although I wasn't blogging back then, so can't reference an article to justify my claim. I have a document in the Burton Group library that I wrote in October 2006 that states: "Watch for integration with IT management and governance: A SOA governance program should be an extension of an enterprise's existing IT governance program. SOA governance processes should blend naturally with traditional SDLC and IT management processes. Registry vendors have yet to deliver integration with configuration management databases (CMDBs) or IT Infrastructure Library (ITIL) systems; however this type of integration should start to appear in 2007." Like Todd, I was a bit overly optimistic. I suspect that lack of standards for CMDB has constrained it growth into a general-purpose configuration management system. No one is yet considering the idea of managing application configuration files (e.g., WAR and EAR config files), much less service configuration files. CMDB is still pretty much limited to managing hardware appliances and core application infrastructure (database, app, and mail servers). It definitely doesn't get into change and configuration management of software componentry. Comments on Michael's points below... On Fri, Oct 10, 2008 at 5:28 AM, Michael Poulin <[EMAIL PROTECTED] com> wrote: > A couple of also interesting (I think) things to add: > > 1) with some knowledge and efforts, a UDDI may be used in the same manner as > LDAP - you will be the master of its meta-data/schema and you will be able > to store all information needed to "to be able to track a fault/problem all > the way from the business process down to a switch, network card or a cpu in > a server and vice versa" via its programmatic (vs. manual) interface The UDDI data model is very extensible. You can use a tModel to represent pretty much any thing or any relationship. So, yes, you could use UDDI to store all information needed to track a fault/problem all the way from the business process to a device or vice versa. But in order for this information to be useful, you would have to standardize the tModels that capture the information. Standardization of management information has been attempted many time before with little success. Think CIM and SML. > > 2) I have not seen or heard of such use of UDDI as I described in 1). > Probably, Anne will correct me here. Nevertheless, the task "to track a > fault/problem all the way from the business process down to a switch, > network card or a cpu in a server and vice versa" is one of the major ones > on the way to the SO model of the enterprise. I know that IBM hasn't > included UDDI support into is service registry/repository and offered an > alternative solution for this I have never seen anyone capture this much information in UDDI. The Systinet-led GIF effort (now led by HP -- see https://h10078. www1.hp.com/ cda/hpms/ display/main/ hpms_content. jsp?zn=bto& cp=1-11-130- 27^2804_4000_100_ _) defined standard tModels to capture service configuration information -- i.e., policies -- in order to enable interoperability among heterogeneous SOA infrastructure components. Quite a few vendors signed up to participate in GIF, including BEA, AmberPoint, Layer 7, Parasoft. At one point they had more than a dozen vendors participating in the effort. But GIF never attempted to capture configuration information down below the service level. GIF is still around, but I get the sense that interoperability among heterogeneous vendor products isn't as high a priority with the vendors as it once was. > > I agree with Kelly Emo on the mentioned statement: SO registry/repository > needs some things that CMDB and UDDI include today, especially, with regard > to governance (policies/procedure s) and operational business service/process > changes. So, I solute to Todd with his Provision-2006. I doubt that CMDB will every be expanded to the point where it supports design time metadata management. There comes a point where you really want to separate design time and runtime information management. SOA repositories should support design-time governance processes. SOA registries typically support runtime governance processes (e.g., information exchange among runtime components). I would still like to see CMDB move up the stack and take over the runtime management role from SOA runtime registries. Integration among all these repositories (automatic propagation of information from one to another as services proceed through their lifecycles) would be really helpful. - Anne > > - Michael > > ----- Original Message ---- > From: Dennis Djenfer <[EMAIL PROTECTED] se> > To: service-orientated- architecture@ yahoogroups. com > Sent: Thursday, October 9, 2008 8:45:51 PM > Subject: Re: [service-orientated -architecture] ITIL & SOA Governance > > The current organization I'm working with are in the progress of upgrading > from HP SOA Systinet 2.x to 3.0, and there has been a lot of talk about how > the information in the register/repository could be integrated with a CMDB. > The IT operation would like to be able to track a fault/problem all the way > from the business process down to a switch, network card or a cpu in a > server and vice versa. I haven't seen a fully automated solution to this > yet, and I haven't had the time to participate in the evaluation of HP SOA > Systinet 3.0, but it was interesting to read this article, where Kelly Emo > from HP says: > > "The new SOA infrastructure component captures more than UDDI information, > it encompasses best practices, CMDB information, and sets the stage for a > wider culture of governance" > > http://it.toolbox. com/blogs/ dana-gardners- briefing- direct/with- > systinet- 30-hp-broadens- soa-governance- role-to-encompas s-services- > lifecycle- business- processes- it-service- management- 27584 > > > // Dennis Djenfer > > > Todd Biske wrote: > > I'm not that familiar with the ITIL v3 processes (just got registered for > some training in November), but from what I understand of IT Service > Management, I see no reason why the techniques of service management > shouldn't be applicable to either IT services (e.g. Build me a new server) > or business services that are supported by It solutions such as a Web > Service. > As for the tooling, I absolutely think that we'll see convergence in the > CMDB space and the Service Registry/Repository space. I've blogged on this > in the past, as early as August of 2006, but my prediction has yet to come > true. I think there's still too big of a gap between the development side > of IT and the operational side of IT to really establish a market for a > converged product. > Blog: > http://www.biske. com/blog/ ?p=64 > -tb > Todd Biske > http://www.biske. com/blog/ > Sent from my iPhone > On Oct 8, 2008, at 11:54 AM, delarco71 <[EMAIL PROTECTED] es> wrote: > > Dear friends, > > How ITIL v3 processes could be applied to Services Lifecycle ... or in > SOA Management area?. > > Is it possible that SOA/ITIL vendors approach in a future a > convergence between the CMDB and Repository products? > > Regards, > > jose > > ____________ _________ _________ __ > > No virus found in this incoming message. > Checked by AVG - http://www.avg. com > Version: 8.0.173 / Virus Database: 270.7.6/1714 - Release Date: 2008-10-08 > 07:01 > > > >
