Welcome!

Open Source Authors: Pat Romanski, Elizabeth White, Liz McMillan, Jnan Dash, Lacey Thoms

Related Topics: Big Data Journal, Java, SOA & WOA, .NET, Virtualization, Cloud Expo

Big Data Journal: Article

The ‘A-ha’ About Active and Inactive Data

Despite advances in technology, the way we manage the growth in files has not changed much in the last decade or two

How many kinds of files are in use in your organization? Fifty? A hundred? More?

Wrong. There are only two kinds of files: active files and inactive files. (I admit it was a trick question.) Users create a file - a spreadsheet, an image, an invoice, etc. As long as the file is being worked with, it's an active file. Users can share, collaborate, and edit the file to their heart's content, but eventually the file is no longer current, and it becomes inactive.

But in this litigious and increasingly regulated world, there is no such thing as deleting a file. You're stuck with that inactive file, for years, if not forever, like your 30-year-old college dropout kid who still lives in the basement.

Despite advances in technology and a proliferation of alternatives, the way we manage the growth in files has not changed much in the last decade or two. We let the volume of files grow until it consumes all of the available resources, and then we throw money at the problem. Or we throw money at someone else who does essentially the same thing, just with cheaper disks.

Years ago, it was not a bad strategy. The majority of all files were active. The cost of having inactive files in the same pool was small. Today, it's an outdated strategy, and it's a costly strategy, one that few CIOs and budget-minded IT administrators feel particularly good about anymore. That's one positive side effect of the global economic recession: operations of all sizes, including Fortune 1000s and large government agencies, have much less tolerance for deferring a problem by throwing money at it.

It frankly makes no sense to spend thousands to millions of dollars on more and more primary storage to store obsolete or redundant files. It makes even less sense when we consider that runaway data growth is something of a myth when it comes to active files: in most organizations; the volume of active files - the files people create and use to do their jobs day in and day out - remains nearly fixed.

How can this be? Consider the way most workplaces operate. Your company pays its staff to work on different things this week than what they worked on last week or last month. Employees create new files for each new project, and their previous ones become inactive. In a typical organization, inactive data is responsible for data growth, not active data, whereas, the number of active files is proportional to the number of employees available to use them. For the most part, in three years, you'll have about the same number of active files as you have today, assuming you have the same headcount. Most companies add a few employees, let a few employees go, generate more files during the busy season, then fewer files during the lean times. In general, however, the number of users and the amount of active files will stay relatively constant. In this post melt-down economy, no one is hiring like there's no tomorrow. In fact, five-figure layoffs are still a common occurrence; they just don't make the news anymore.

If you've been in business for at least a few years, it's safe to assume two-thirds of your files are inactive. What's more, once saved to the network, 80 percent of your files will never even be looked at again.

Note that "looking at" an old file is not the same as "working on" an old file. An inactive file may be requested for research or reference purposes, but it won't be edited. When it reaches that stage of inactivity, it's in essence a read-only file. This means your inactive file data should be managed as a read-only archive - since the odds bet is read never, and the next best bet is read-only.

With these considerations in mind, purchasing additional primary storage to manage inactive file growth is unjustifiable, especially considering that storage operating costs can be five times the initial acquisition costs.

The growth of inactive files has an additional administrative cost when one adds on the constant cycle of backups. It's not unusual for an organization performing a standard weekly full backup, plus nightly incremental backups, to have 15, 20, even 25 copies of file data at any time. One of our customers, a multinational banking and financial services firm, was maintaining 28 copies of each file until we came in to provide relief.

Employing the same old backup tactics with online storage saves money on hardware, but does not address the core problem of file proliferation. Transferring inactive files to less costly archival storage media works well on paper, but not in practice. Both these methods suffer from inefficiencies and insecurities. Physically moving a petabyte of data can take months, and finding a file takes arduously long (if it's ever found at all; more than 60 percent of the time, it's not). This can spell the end of the business when a federal court judge comes around asking the company to produce a few thousand documents.

The real savings kick in when the typical file backup strategy is scrapped in favor of tiering, which is based on an online archive for inactive files. The key to achieving this is controlling and managing how and when files are moved - in other words, when a file becomes inactive.

The goal is to categorize data, properly manage it, and move the right data to the second tier to reduce costs, address compliance issues, and meet electronic discovery requirements. Granted, the cost, time, and complexity involved in scanning the entire file system and moving data to an archive can be onerous. However, event- and policy-driven software that automates these processes based on age, type, size or access frequency makes this a much smaller chore.

With this method, it's not unusual to see a two-thirds reduction in the cost of storing and protecting inactive files in the first year alone, and the savings accumulate as the inactive file load grows. One manufacturing customer calculated their first-year savings as 82 percent. Even if the initial costs of this tiered approach are twice as much as storage hardware (which seems unlikely), there would still be a savings as time goes by, since operating costs always exceed acquisition costs.

This second tier for inactive data can be maintained as a read-only archive, since it's very unlikely these files will be changed again. For example, an engineering firm may want to view the architectural plans from a project they did two years ago, but it's unlikely they would make any edits on that original file. This can deliver additional savings over time.

If no other message resonates, we hope we've at least established that traditional methods of accommodating data growth do not make sense when it's inactive, infrequently accessed, obsolete files that are typically responsible for that growth. Expensive primary storage, and the Sisyphean task of regular backups, will not solve the problem of inactive files.

More Stories By Bruce Backa

Bruce Backa, CEO of NTP Software®, a leader in File Data Management, is recognized as a Technology Pioneer by the World Economic Forum. Prior to founding NTP Software nearly 20 years ago, he held CTO positions at the international insurance and financial services firm AIG and the American Stock Exchange. He holds several U.S. and international patents related to data management, and was awarded the Kemeny Prize in Computing from Dartmouth College.

Comments (0)

Share your thoughts on this story.

Add your comment
You must be signed in to add a comment. Sign-in | Register

In accordance with our Comment Policy, we encourage comments that are on topic, relevant and to-the-point. We will remove comments that include profanity, personal attacks, racial slurs, threats of violence, or other inappropriate material that violates our Terms and Conditions, and will block users who make repeated violations. We ask all readers to expect diversity of opinion and to treat one another with dignity and respect.