Processor ® Free Subscription
Used HP, Used IBM, Used Compaq, Used Cisco, Used Sun
Home |  Register |  Contact Us   
This Week's Issue
Browse All Issues
Search All Articles
Product News & Information
Company
News & Information
General Feature Articles
News
Opinions



How To Email This
Print This
View My Personal Library

General Information Add To My Personal Library
October 9, 2009 • Vol.31 Issue 25
Page(s) 30-31 in print issue

Get Started With Capacity Management
The Alternative To Throwing Hardware At The Problem

Key Points

• Capacity management solves such problems as how much growth existing systems can cope with before they break, how much hardware will have to be purchased, when to cope with anticipated growth, and what is behind a performance bottleneck.

• Learn before you leap. Get clued in on capacity management via education and training.

• Start with the low-hanging fruit. Don’t try to solve all organizational issues in your first week.

In some circles, capacity management is regarded as something of an esoteric art. Yet when you look at the list of proponents, such as Verizon, Visa, HP, IBM, and Oracle, you get the idea there must be something to it.

OK, so what is it? A group of standards and processes collectively known as the ITIL (IT Infrastructure Library) defines it as a group of work processes associated with the provisioning and management of IT resources.

“Capacity management is the proactive task of forecasting, anticipating, and adjusting to growth or other changes to minimize or eliminate the impact on the organization,” says Steve Wong, VP of marketing for ClearSight Networks (www.clearsightnet.com).

What does it accomplish? According to G. Jay Lipovich, director of product management for Data and Performance Management at BMC Software (www.bmc.com), it tackles such things as how much growth existing systems can cope with before they break, how much hardware will have to be purchased, when to cope with anticipated growth, and what is behind a performance bottleneck.

Accomplishing those results, though, is not a plug-and-play matter. It takes hard work based on really knowing what you are doing. Education, therefore, is the first step to be taken in any capacity management journey.

“Awareness and education are the best ways to ensure a successful implementation,” says Wong. “Speak to your industry peers at other companies that have successfully rolled out capacity management programs. In addition to helping you navigate through the process of tools selection and data center deployment, they may be able to help you avoid the common pitfalls and mistakes that are often made—because they have made them.”

Once you are clued in on the subject, BMC’s Lipovich says that it is vital to gain management buy-in before commencing.

“Gaining senior management commitment to the value of capacity management is key,” says Lipovich. “There must be organizational commitment to the efforts, because cooperation is needed from the lines of business, development, IT operations, and finance to achieve results with the highest value to the enterprise.”

Steps To Take

Assuming that a basic understanding of capacity management exists and that management buy-in has been obtained, the next steps include the following.

Staffing. In smaller sites, a few generalists do everything. One of them can be assigned the capacity management hat. Larger sites demand several people to look after the various aspects of capacity management—some to plan capacity for the future and others to manage the immediate performance concerns of applications.

Ron Potter, manager of best practices at capacity management vendor TeamQuest (www.teamquest.com), says he once thought mathematicians would do well in capacity planning due to the formulas and statistics involved. If such people lack IT experience, though, it’s impossible for them to judge the reasonableness of the conclusions.

“Only experience and intuition can accurately judge whether a result is reasonable,” says Potter. “For the most part, I have found that system programmers and system administrators make the best capacity planners, especially those with considerable operational experience.”

Training. Clearly, capacity management is an area where training is required. Plenty of training courses are available from ITIL-related organizations. Groups such as CMG (Computer Measurement Group) organize courses, seminars, and conferences. Vendors also offer training and conferences on capacity management basics and the use of their own tools. There is no doubt that investing in capacity management and process-based training will pay big dividends in the long run. Make sure, then, that whoever is assigned to these duties is supported in his or her efforts to learn the basics of the subject. Sending the staff to an annual conference is another good way of staying up-to-date.

Processes. ITIL is a popular approach to IT processes. There are others which go under a variety of names.

“Processes ensure you come up with the same answer every time you analyze the same pieces of data,” says Potter. “Meaningful, repeatable, and reliable results generate credibility, something capacity management needs to survive.”

Sample processes can be obtained in ITIL books or from consultancies and vendors.

Data. The type and amount of data you need to collect depends on the desired goals. For servers, collect data such as processor busy, disk read/writes, memory statistics, job/process execution statistics, and database-related statistics. Networking capacity managers, on the other hand, commonly look at bandwidth, transmission delay, router, firewall, and packet statistics.

“The types of data depend on capacity management goals,” says Potter. “In most organizations, too much data is collected initially, then reduced and refined as the capacity management organization matures.”

Whatever data is gathered, it should be captured in the same intervals across all infrastructure components. Otherwise, it is difficult to analyze the interactions between a transaction server and a database server if one is collecting at one-minute intervals and the other at five-minute intervals.

Tools. Tools are a core element of any capacity management department. OSes such as Windows, Linux, and Unix come equipped with various tools to measure and trend performance data. Unix, for instance, offers VMstat, TOP, and other freeware to record basic statistics of system performance. For some organizations, using these along with spreadsheets may be enough. Others may need to make a more comprehensive address to the subject using products from capacity management vendors.

“Contact at least three vendors and ask them to show you case studies of how companies similar to yours have deployed their tools for the purposes of capacity and performance management,” says Wong.

Initial Success

Some who are enthusiastic about capacity management begin by attempting to solve all organizational concerns in one fell swoop, which is a big mistake.

“Do not try to ‘boil the ocean’ in initial efforts for capacity management,” says Lipovich. “Begin with a small focus, demonstrate the value of the effort, then develop a larger scope.”

That might mean concentrating effort on mission-critical resources or poorly performing applications before your enterprise as a whole. However, the quick results that stem from capacity management can transform the technology from a dark art into the flavor of the month.

by Drew Robb

Top Tips

• Select a capacity management database capable of maintaining metrics from all OSes and applications in use, not just one platform.

• Software costs range from $20,000 to more than $1 million. Start small, and see how it goes. If the ROI is there, expand the scale of your deployment.

• Educate yourself using the Internet, by reading the capacity management chapters of the ITIL handbooks, or by joining the CMG (Computer Measurement Group), which has an annual conference and local chapters for capacity managers.

• Don’t get too bogged down in the details. Because capacity planners deal with averages, working at too fine a level of detail can bog down analyses and has little or no improvement on the precision of the work.
Share This Article:    del.icio.us: Get Started With Capacity Management     digg: Get Started With Capacity Management     reddit: Get Started With Capacity Management

 

Home     Copyright & Legal Notice     Privacy Policy     Site Map     Contact Us

Search results delivered by the Troika® system.

Copyright © by Sandhills Publishing Company 2010. All rights reserved.