brian.thomson@mahc.ca

Altiris CMDB Solution 6.5

I'm intrigues by some of the resource types available in CMDB especially the capability to document dependencies with the 'datacenter' types and the 'service types'.

I think I have intent behind the various datacenter resource types figured out.

I am trying to understand what the intent was behined the 'service types' which include service and system resources. As always documentation is pretty thin on how these can be used. In the documentation they provide an example of an 'email service' type.

It is difficult to understand the difference between "Dependent CIs" and "Supporting Systems". What is the difference between a "system" and a "service" .

Does anyone have experience with this part of CMDB that they would like to share. Are there any Altiris whitepapers that do a better job of explaining this?

Thanks in advance,
Brian

gdellaccio

Hi Brian,

We use CMDB 6.5, and you are correct that it can be confusing and not well documented.

The difference between 'services' and 'systems' are based on ITIL & ITSM definitions. I recommend reading up on ITIL & ITSM to fully understand what Altiris is trying to do with CMDB.

In a nutshell, the concept of "Service" in an IT sense has a distinct operational connotation, like e-mail. E-mail is service provided by your company.

A 'System' can be the way a service is delivered. A system can be an Exchange server, or series of servers. It can also include things like the routers and switches needed to access the 'system'. Each one of these items (CIs) is dependent on something further up the chain. For example if a switch between your Exchange Server and your users goes down, it does not matter that the Exchange Server is working fine. The Service of e-mail is “dependent” on the systems, in this casr switch, related so the services can be delivered.

Each item in system, computer, user, server, etc. can be a CI. You can run an Altiris Agent on it and that makes it an automatic 'Computer' CI, or you can create a static instance of that computer/server and make it 'datacenter type'.

The reasoning behind this would be like in this example:

An IT manager wants to shut down server name ‘1234’. Unbeknownst to this manager, this server provides e-mail to one small remote office only, but on the surface does not appear to do much. If shut down, the service of e-mail cannot be delivered to that office.

By setting up the parent-child or 'Dependent CI' relationships in CMDB, you (or your company) would immediately be able to look at the Resource Association Diagram (RAD) and find out that shutting down that server, while it seemed minor would actually cause an e-mail outage to this one site, or the groups of users at that location.

The trick is finding out this information and bringing it into the CMDB, and having a framework like Change Management meeting to discuss how different systems relate to each other. Odds are, they are more dependent on each more than IT thinks.

This feature requires a lot of planning, but can be put in place over time. This pieces not needed to use most of the functions of Asset/CMDB. It is more like Advanced CMDB.

There is TONS of information on IT Service Mgt, so I’ll stop there.

Hope that helps