Your Data Protection Plan: Where It Should Be and How To Get It There

By Julie Lugar - Reprinted from Contingency Planning & Management Magazine

Business growth can require changes in data backup and recovery management techniques, especially if existing hardware and networking schemes are expanded or redefined.

The best management plans handle data coherently across the entire system. The worst approach, and the most costly to the bottom line, is trying to manage data on a server-by-server basis. When information means money, protecting data in a shoddy manner is as devastating as not having a handle on the company's finances. In today's world, data is inextricably linked to financial profits.

"We're trying to help a company right now that is struggling with their antiquated network systems," says Nick Cornacchia, president of Yosemite Technologies, a provider of enterprise-class storage management software and data recovery solutions. "They have lost a great deal of money and their stock has dropped in value because they don't have a handle on managing their data. The system in a company may be running perfectly every day, and so they think that 'if it isn't broken, don't fix it,' but in the fast-moving world of today, productivity is where it's at."

The recent changes in e-technologies and e-business models means rethinking and readjusting the methodologies used to back up, protect, and secure mission-critical data for recovery in case of a disaster. Networking environments have changed radically over the past few years. The move to data warehousing, data mining, and e-commerce business-to-business and business-to-customer transactions has added administrative burdens to already overworked IT staffs. Networks are more heterogeneous than ever before as they have grown to serve a multitude of sites that span a country or cross borders. The requirements of 24/7 data availability, and the shortage of personnel for around-the-clock management has helped lead to remotely managed "lights-out" sites. As any administrator knows, this kind of environment is a true management challenge any way it's approached, and finding the correct formula for the solution isn't the cut and dried implementation that works for simple LAN networks.

"The truth today is that a lot of different variables have to be analyzed and incorporated into a successful business," says Cornacchia. "Data backup is the most important thing to be done if the business is doing any business at all. The data on those servers is critical. If it's lost, the business could fail."

Finding the correct formula for your new data recovery paradigm isn't a difficult process.

Analyze Current Needs

Qualify the current network hardware and software, particularly in terms of how much data will be held on the servers, the amount of data that is handled on a daily basis, the devices already in place to perform data backup and recovery, and the software running those devices. Analyze how the current network configuration is administered and managed: how many servers, their location, where they are administered from, where and how the data is backed up. Look at the current disaster recovery plan to see how it fits in.

"There should be schemes in place to insure that the data can be restored very quickly if something happens," says Cornacchia. "If there isn't a disaster recovery scheme on the system, it should be added. Good disaster recovery software will allow you to restore all your data in a matter of minutes."

A backup solution that requires an administrator to run daily and weekly backups manually from each server will be adequate for a small business environment with one or two servers and minimal amounts of data. In an enterprise environment, such a solution will fail rapidly under the pressures of business growth, and is unthinkable in any kind of a distributed environment that must have some form of remote administration to be managed effectively. The analysis, if properly done, will reveal how effective the current backup scheme is in supporting a business' current network.

Project Future Growth

The next step is an analysis of future needs. The current business plan should provide some kind of guideline as to the expected growth over the next year, and a projected guideline for the following years. Comparing current levels versus the projected/desired growth will give a baseline on the future needs that must be planned for, such as adding additional employees, functions, or business processes. Compare the current levels of data with what might be expected in the growth scheme of the business. Analyze disaster recovery plans, and how often data is updated or exchanged if off-site storage is utilized. Review the current management techniques to decide if this growth will dictate new management techniques, and if there are problems that require solutions. The biggest change may be to redesign data backup management techniques to accommodate business growth.

"We talk to businesses every day that have backup hardware and software that is six or eight years older than the system they are constantly upgrading and can't figure out why they can't use it effectively anymore," says Cornacchia. "There is no one-time investment in data backup solutions, yet some companies don't analyze what their maintenance and upgrade needs or costs will be when they buy their system initially."

Investigate Solutions

An ideal data backup solution will allow an administrator to go beyond yesterday's solution of a machine-to-machine model of data backup to an enterprise-wide solution for data backup. Any software solution for data backup and recovery management should have the following features:

• It should work over an entire network/server setup from a central location and be operating system-independent.
• It should be flexible enough not to impose constraints on the development of a backup plan, and easily adapted to future growth.
• Data can be sent to the best backup device available, whether that is across the network or the local bus of the server. The criteria for deciding this is network speed and bandwidth requirements, the amount of time available for backups, and the availability and capacity of the backup devices.
• It should be able to run concurrent backups from different machines. Concurrent backups are more efficient and minimize server downtime, thus shortening the time span in which data is temporarily unavailable.
• It should have automatic task scheduling that can be managed from a central location to eliminate the need for an administrator to put "hands on" each machine to perform backups. This is the only solution that can work for a "lights-out" environment and still ensure that the data is properly protected.
• It should be able to utilize any server as the central administrative point for managing data backups. This gives the ultimate flexibility in designing a backup solution to match current network configurations and future growth.
• It should incorporate "one-button disaster recovery." The faster a disaster can be remedied, the less damage a business will suffer.
• It should be able to automatically move the data to a different backup device if the primary backup device is unavailable or full.
• It should track and manage the location of the backed-up data automatically, so that the administrator knows exactly what is on each piece of media.
• It should be robust enough to handle large amounts of data and multiple machines and allow the administrator full control over all the data and each machine.

Implement the Plan

Once a coherent plan has been developed to accommodate the current business needs with room for future growth, it should be implemented as quickly as possible. Careful attention should be given to each stage of implementation to assure that the solution is workable. Practice disaster recovery restores should be tested with new software and hardware to make sure it will work when needed. Waiting until the disaster happens to find out if your disaster recovery solution actually works is taking an unnecessary risk with the business assets. An administrator needs to know that the backup system in place will work each and every time. Initiate the backup plan and monitor it to make sure that the backups are performed correctly and everything is in place and functioning properly.

"One of the biggest problems for companies is to have a backup system and never use it, or never use it properly," says Cornacchia. "Maybe the system they have is too slow because they bought an inexpensive tape drive that doesn't run fast enough, or maybe they have to go down and run the backups from each machine directly and they don't have the time. Having the system isn't enough. It's got to be used, and used properly, if there is to be any benefit."

Review Management Plan

Any data backup plan needs to be reviewed annually or semiannually to make sure that it is on track with the business' actual growth, is delivering the kind of protection the business needs, and is not breaking the bank to administer and maintain. This review process can be naturally fit into most business' financial scheduling and budget planning periods.

"Networks are a fast-moving arena nowadays, with all the changes to operating systems, bigger drives, and complex data storage solutions coming to the market," says Carnacchio. "Systems change over time and a company needs to change with them so that they never end up with a backup solution that is antiquated. Cost for upgrades are small compared to neglected backup strategies that could cause loss of data and necessitate the high cost of recovery."

<< Back