ITSM

IT Service management (ITSM) primer <<previous|next>>

Data Center News:

ITIL worship: The new set of best practices

By Luke Meredith, News Writer

13 Jun 2004 | SearchDataCenter.com

NEW YORK -- Columbia University's Anthony Johnson has quite the challenge on his hands.

Johnson is a systems administrator working to put together an IT staff for the Marcus Langseth, Columbia's ocean research vessel that is set to launch in 2006. But finding data center staff for a ship isn't as easy as hiring a few techies for a land-borne server farm. You can't keep data center employees on the Atlantic Ocean year round, making constant turnover a certainty.

One of Johnson's main tasks when Marcus Langseth hits the high seas will be keeping the constant stream of new hires up to date on standard practices. So for him, the best practices models offered by the IT Infrastructure Library (ITIL) stretch past annoying compliance concerns.

ITIL, specifically change management practices, scratches Johnson just where he itches.

The larger you are, the more important it is because of the more pieces you have at stake.
Millard Jackson
data center operations senior directorSiemens

"We're in a unique environment … you don't have people who are on 12 months a year. You have people on for a few months who then leave," Johnson said. "Change management is something I'm very interested in … keeping continuity while you're going through people like that is certainly very attractive."

Whether it's service-level availability or change management, ITIL -- a cohesive set of best practices supported by a comprehensive qualifications scheme, accredited training organizations, and implementation and assessment tools -- is generally considered the most widely accepted approach to IT service management.

Launched in the U.K. in 1989, ITIL is based on service management as a documented framework of best practices for managing IT infrastructure.

There's a lot to consider -- and much of it is overkill, especially for small companies looking to do a little more than keep a few racks in line -- but attendees who attended a session entitled "Implementing the IT Infrastructure Library (ITIL) in Today's Data Center" at last week's Data Center Decisions conference agreed that ITIL has a place in the modern data center and is gaining steam as a legitimate tool for companies seeking to keep their IT systems in line.

Kevin Congreve, a data center manager for Lawrence, Kan.-based Mackenzie and Co. Inc., a management consulting firm that operates out of the U.S., Germany and Australia, sees ITIL as a critical tool for establishing basic best practices guidelines between his data center and those overseas.

Congreve and his team got certified last year, and while he said there's little need for smaller organizations to become ITIL disciples, much of what it offers is simple common sense for anyone looking for a streamlined server farm.

"From an overall operations perspective, it's pretty important … making sure that changes that occur on your network, [such as] adding servers and hardware and upgrading operating systems [are accounted for]. If you don't know what's going on, [you] are trying to find out a problem that you could have easily have found, so there are a lot of pieces that are important to an operations group and specifically the data center."

Of course, the more heterogeneous your data center, the more ITIL can help. Millard Jackson, Siemens' data center operations senior director, works in a very large data center featuring around 1,000 servers -- most of which are Intel boxes.

But Seimens also deploys seven mainframes with a total of 16,000 MIPS of mainframe power. As a data center manager with experience keeping a handle on mid-range boxes as well as servers, Jackson sees ITIL as a model more suited to Intel-based organizations, simply because of the nature of a server farm chugging along with racks and racks of small servers.

"The larger you are, the more important it is because of the more pieces you have at stake," Jackson said. "It's more important for the server world because it's much more complicated. The mainframe world is much easier to track because of its nature. The open systems world is a nightmare … if you don't have these policies and procedures in place you're not going to know where anything is."