Comments on: What are the Six Functional Areas of Data Center Automation

By: Ryan Shopp

Ryan Shopp — Fri, 14 Dec 2007 23:05:43 +0000

Great points, this is a work in progress and i really encourage/enjoy others input & ideas. I’m looking forward to maturing this blueprint over time.
Storage, yep – that should be a data center infrastructure category.
Virtualization as a layer above the data center infrastructure stacks needs some pondering…i see where your saying in the future it may be an abstraction layer in between today’s data center infrastructure categories and data center automation functions.
Great questions about the hierarchical relationship…i don’t believe it evolves that way, many times the exact opposite in my experience. The point I’m attempting to highlight is the lower in the stack means you’re tied closer to each data center infrastructure category. This requires more domain knowledge for that part of the infrastructure. Example: The expertise level to understand how to make a specific setting change is deeper then understanding a cpu performance bottleneck which is deeper then knowing if something is available. That’s what I was going for in this first version of the graphic conceptually.
Help Desk & CMDB are in there to the top and sides currently under different names…help desk (e.g., manual tasks) and CMDB (e.g., resource reconciliation)
Analytics is something I need to put more thought into as it may be a functional group all to itself that spans the other 4 functional automation areas (e.g. put it off to the the right) or something to considering as part of the dashboard which I’m not currently putting in scope of DCA in this blueprint.
Lot of great feedback. Thanks! I’m planning to take another stab at the graphic early next week.

By: RStreu

RStreu — Fri, 14 Dec 2007 20:20:01 +0000

Hi Ryan,

This is an interesting topic. I would add Storage as a distinct Infrastructure at this point. Storage, especially Virtual Storage, will become part of the fabric of IT. All Services and applications will rely on the storage devices.

Another interesting component is Virtualization as a technology stack at all layers: Storage, Network, OS, and Application. As the the datacenter becomes more dynamic, there will be increasing complexites introduced by these abstractions. Where does that fit in the model?

Also, did you intend to convey a hierarchic relationship between configuration – security – performance – availability? Do you believe it moves in that direction? My thought is configurations provide services. The different configurations result in different resultant characteristics and behaviors with metrics such as how secure, how compliant, how available, etc.

Where do you put an ITIL type function? Where’s the Service Desk? The CMDB? This is where the Processes shoudl start and end with interaction with the other tools/functions.

Also, true Analytics would be looking down at or across all of this to help steer IT in a process of continual improvement across people, process and technology.

This is good stuff. Keep it going!

By: MBro

MBro — Fri, 07 Dec 2007 19:09:19 +0000

Hi Ryan,

That’s quite the ambitious entry. I’m interested to see how your reference model develops. Let me know if there’s anything I can do to help.

- Mike