Infrastructure Consolidation    

By Susan Ibarra 

The final issue of a 3 Part Series

 Back to Articles 

Many of the folks at Innovative Systems Design came out of AT & T.  David Nocera, Vice President of Innovative Systems Design explains, “As Chief Architect for AT & T Long Distance, I had an infrastructure of 475 Unix and NT servers, located across 35 call centers and 2 data centers.”  He continues, “When you manage something that vast you have to come up with procedures and methodologies for doing it in an efficient way.  Otherwise, you have 10’s of people managing infrastructure and your operational costs are driven even higher.”  Some of what Mr. Nocera and his associates have done at Innovative Systems Design is to take their methodologies and turn them into software.

Mr. Nocera continues, “There are many different questions at different layers of the infrastructure.  Regarding the analogy of a city to infrastructure (see Vol. IV-A), one infrastructure equivalent would be the storage area networks.  This would be the SAN, NAS and your network appliance.  This is the storage layer or tier.  You also have server tiers, which could be Unix infrastructure or NT.  It could be multi tiered which would have a presentation layer happening at the web tier.  You could have applications and business logic at the application tier and storage happening at the database tier of a server solution.”

“Then you have the network and connectivity between the LAN and the WAN.  It’s a good analogy to use city utilities as a metaphor for the utilities that are available in an infrastructure.”

According to the professionals at Innovative Systems Design, the latest way to manage large infrastructure is known as virtualization.  Virtualization is creating logical representations of physical infrastructure.  For example, storage.

Physically, storage is located in lots of different places within an infrastructure.  Storage could be disks, tape backups or online backups of your database.

When you put a layer of storage virtualization between your application server and your storage, you can now dynamically control and manage the storage independently of the vendor or type of storage that exists underneath.  Often, utilizing a storage area network solution is a far more flexible environment with greater cost savings.

A major issue: Infrastructure Consolidation. 

Infolutions:  Dave, how do you reduce the cost of a company’s complex infrastructure and maintain technology at the least possible cost? 

Nocera:  “Strategies and methodologies that my team has implemented to save companies money were generalized into “Best Practices” and software.  As a result, companies in various industries use our Best Practices and tools to identify opportunities to save money.”

Infolutions:  Opportunities?

Nocera: “There are always opportunities to be found in large infrastructure.  Very often, software squatters are uncovered.  Once, a long time ago, software was purchased.  Now, many lifecycles later, that software is no longer being used.  Somewhere in the company, somebody is still holding on to it, just in case they need it . . .  In the meantime, someone else in accounting is paying maintenance costs on that software thinking it’s a legitimate piece of software bringing value to the bottom line.  But it’s not because it’s not being used!  That is an example of an untapped asset within an infrastructure.”

How can that happen?  Well, in large infrastructure there are always problems like that because the way business works, it becomes complicated.  Quite often, the person who bought the software is no longer there.  The new person coming in to fill that roll doesn’t know if the software is going to be used or if the purchase was for future planning purposes.

Double Troubles, Toil and Billing

Nocera:  “When hardware is ordered there are usually two addresses.  The address of the one who ordered the hardware and the ship to address.  Very frequently the one who ordered the hardware gets the bill . . .and it sits in a file somewhere.  So when the bill does not get paid on time, a duplicate bill is sent to the shipping address.  Eventually, both invoices get forwarded to the Purchasing Department who winds up paying both invoices!”

As sloppy as this is, it happens all the time.  It’s really an issue of Infastructure Intelligence©.  Simply put, — How to manage your assets more efficiently so your technology investments are aligned with the value they bring to the business (same money).  This is what is missing in business today.

Infolutions:  How can these things happen?  Isn’t there documentation so Accounts Payable knows what to pay? 

Nocera:  “Unfortunately, with large infrastructure another inherent problem is that things get bundled up and rolled up until the people can’t see the forest thru the trees.”

“You see, once upon a time in the year 1, there was a project and it had documentation.  So in this year of 1, the project is funded and gets deployed.  Then the people are happily ever after sent off to new projects.  So now there is no one left to tell.  Oh, there might be some core group of people that are responsible for the project at this point.  However, they are nowhere near the level of expertise, they don’t have the vision, they may not know anything.  They could even be outsourced!   They don’t understand the big picture—they’re only looking at the parts!”

“Someone could come up to them and say, “…your using BMC Control”.  The guy might even be turning in on once in a while, but it is not doing what it’s supposed to be doing and it is NOT bringing any value.”

Typically, in the year 1, the documentation starts off quite accurately and there is a lot of architectural backup.  So in a situation that an application is written to use 5 servers, in year 1 it needed 5 servers.  But applications have and go thru a lifecycle.  New applications come along and replace it.  So over time, the functionality of the application becomes outdated and inefficient.  However, the software that is being licensed for that application to function continues to be paid for, under the assumption that it still is viable and efficient.  This continues because no one peels back the onion to that level of detail.  Those are the types of large holes that Mr. Nocera often finds.  This style of infrastructure consolidation requires total understanding of how to save money on existing assets.

Another tactical approach to show and tell companies exactly just what it is costing them in extra time, money and assets is something called, Modeling.

Mr. Nocera explains, “Many times companies will ask for hardware maintenance contracts for a 4 hour turn around time.  But they are paying for a 2 hour turn around time.  As a result these companies wind up over paying for the infrastructure because they’re paying for a faster turn around time.”

He continues, “When we model an infrastructure, we are in effect, building a mathematical model that takes into consideration business requirements and current technology, then predicts future technology utilization.  By modeling we can better align the technology with the needs of the business, showing the costs and risks of adjusting these maintenance contracts and determine what we are getting for that extra two hours.” 

When any company’s infrastructure is designed properly, some equipment failure or downtime will usually not affect company operations or put the company down.  This especially applies to medium and large companies. 

If you are a large company, you are defined as large by having more than 50 servers.  So if some of your servers go down, the multitude of remaining servers can pick up the slack without affecting your business performance, while you wait for the downed equipment to be checked out and restored.

On the other hand, a small company running along on a few servers may find it difficult to keep their company from actually going down when they experience equipment failure.  So a small company will want the technician out there ASAP.

A large company might select the Platinum contract over the Silver contract.  Unfortunately, these companies thought they were buying a higher level of insurance. 

According to Mr. Nocera, “The outages are so infrequent that when you feed the data into a model, the difference doesn’t really amount to that much.  It might be the difference of  99.999 % availability VS. 99.9% availability.  And that might boil down to 5 minutes VS. 45 minutes of equipment downtime a year not company downtime.

Infolutions: How do you fix these problems?

Nocera:  “The opportunities to align business decisions and technology all revolve around having the high quality data. Frequently that data is unavailable for reporting and analysis. Our patent pending technology pulls data from sources and stores it into a single optimized database for analysis and optimization.  So the answer is good data drives good decisions.”

That translates into how a company can save millions of dollars!!!

Do you have a topic that you’d like to read about or need written?  We’d like to know.

Do you have a question about your business?  We can help. 

Call us at 908.322.4020  

© 2001-2014  A Susan Ibarra Publication, All Rights Reserved

Back to Top                            Back to Articles