Holding employees and departments accountable for their use of increasingly precious storage capacity is beginning to gain favor as an efficient approach to storage management. But holding the line on disk usage is an ongoing challenge. Most IT organizations simply expand storage capacity to address demand. Although storage networking (storage area network or network-attached storage) is a good approach to channeling the relentless onrush of data, it exacts a significant toll on administration resources and IT budgets. This is one reason why the concept of “chargeback” is beginning to gain acceptance in organizations seeking to reign in storage use and associated management headaches and costs.
The idea of chargeback comes from mainframes, where computer resources were centralized and expensive to use. Then, most IT departments operated as service providers, making system and resource accounting imperative. Users were charged per CPU cycle, memory page, disk byte, tape inch, etc.
With Windows NT and the advent of client-server computing, cost allocation received little attention. Moreover, NT lacks sophisticated resource accounting functionality. But with NT clusters and parallel processors that function much like mainframes, and with new technology such as SANs, chargeback is making a comeback.
Over the past several years, a growing number of enterprises have implemented usage-based cost allocation and chargeback systems for a wide variety of IT resources and subsystems, including telecommunications, e-mail systems, and external systems such as help desks. Companies can “charge back” costs to user departments, individuals, or customers, encouraging efficient and careful use of resources and, therefore, preserving investment in IT infrastructure.
The idea behind chargeback as applied to storage use is simple: Make people pay for the storage resources they use. Whether it is configured to charge departments for increments of storage capacity above a certain threshold or places a premium on the types of data stored, a chargeback approach holds users directly accountable for their storage. In essence, chargeback is a means of automating the enforcement of storage policies, providing users with strong incentives to manage their resources intelligently and economically.
How It Works
Storage chargeback software contains three basic components: a scanner that collects usage data, a database that stores it, and an engine that can perform simple calculations according to administrator-defined billing rules.
Data scanning-Running a data scan details the files and folders on each host containing user data (typically in a home directory structure). The scan job should determine how much data is owned by each user account that has a home directory on that host. Many storage resource management (SRM) products already keep this data updated in memory, in which case running a data scan job is not necessary.
Data collection-The next step is to decide what kinds of data to export to the database. Once you have configured the database export, the data is available for use in the chargeback software.
Chargeback engine-While chargeback can be managed by in-house applications, specialized chargeback software is becoming increasingly popular. Chargeback software can help companies calculate the underlying costs of IT resources or help calculate how much users should be charged based on the amount of resources they use, while also billing end users.
Once you’ve exported your data and it’s available to your chargeback software, the next step is mapping the fields before it can perform cost allocation and chargeback. Sample fields might include
- Date (or date range) that the disk resources were consumed;
- Identifier fields (i.e., Quota, Object, Server Name, and User), which allow you to map to an account code hierarchy or to drill down to a user or folder name;
- Account code to identify the account;
- Resource (i.e., Money or Units). If this field contains Units such as disk megabytes, the chargeback software would multiply this field by a value contained in its rate table. If this field contained a monetary value, then the value would be aggregated and shown as a charge on the invoice; and
- Rate code, a user-defined eight-character code to identify the resource.
Once these fields are mapped, you can create records for all types of chargeback scenarios, which are then sent to processing programs.
By selecting different rules from lists, you can design chargeback policies such as “The first 100MB are free; beyond that point, calculate a cost of $1 a month per 100MB.” You can even adopt different rules for different users or departments. The application will calculate the cost according to your specification, and the resulting cost and other data can either be printed (for invoicing) or exported to other applications, databases, etc.
You can charge for usage per user, group, or department. Either the administrator or authorized staff members can run the chargeback jobs.
Chargeback is a means of urging employees (or customers) to police their use of storage resources. By affixing a price tag to increments of capacity usage, storage is no longer a commodity; it’s a valuable resource to be used judiciously. Casual data hoarding now comes at a price.
In addition, chargeback facilitates the intelligent and economical allocation of storage resources. By assigning prices according to resource value, users are apt to store data where it is most cost-effective, which ensures that the more expensive storage options are used appropriately.
A chargeback can be an integral component of an end-to-end SRM solution and gives organizations the flexibility to implement a comprehensive SRM solution from the ground up or to extend their existing storage management system.
With chargeback, users across the organization know the price and value of the storage resources they have at their disposal. By holding users accountable, chargeback provides a powerful economic incentive to discipline runaway data accumulation and eliminate the extravagant use of resources.