Planning for Business Continuity

Posted on April 01, 1998

RssImageAltText

Planning for Business Continuity

Providing continuous data access goes beyond disaster recovery and requires a solid business continuity plan.

Ronald Koshko

"Neither rain nor sleet nor snow shall keep us from our appointed rounds." Many IT managers could relate to this mantra of the US Postal service if only "nor sleep" were added to the equation.

The drive for 7x24x365 access to information, regardless of planned or unplanned obstacles, is rapidly becoming top priority in most IT operations. The focus has shifted from disaster recovery, or how long it takes to get a company back on-line after a disaster, to business continuity.

Business continuity is the process of ensuring a company`s revenue stream by providing optimal data access and use. Both the definition and the process contain elements of more commonly used terms, such as high availability, mission critical, fault tolerant, and continuous availability. But for today`s information-dependent businesses, business continuity equals revenue continuity.

Historically, standard disaster recovery methods have focused on the ability to successfully copy data onto tapes, catalog the tapes, and archive them for later use. However, with the growth of information far outpacing the technology curve of tape copy speed, IT managers find themselves continuously faced with a dilemma: which data to back up and how often?

Backup technology has certainly advanced, but no matter how fast it becomes, the time it takes to restore and put a business back in full operation will remain under close scrutiny. IT managers know that every minute spent "recovering from a disaster" results in both lost revenue and opportunity.

The demand for information access may be coming from inside and outside corporations, but the requirement is the same. Users and customers don`t want to hear if an outage was caused by a server failure, planned maintenance, a system upgrade, a tornado, or El Nino`s latest weather tantrum. They simply want continuous, unrestricted access to data.

IS managers are looking at their IT infrastructures--specifically at their information storage systems--to cope with these demands. In the past few years, storage has shed its image as a simple piece of add-on hardware. Enterprise storage has become a critical element in all successful IT plans, and its impact is being felt throughout the data center. IS managers are now looking to enterprise storage solutions to address the demands of business continuity.

A Bulletproof Plan

In the course of planning and implementing a successful business continuity plan (BCP), the CIO, CFO, and data center manager need to make sure the strategy continues to support the organization`s main business interests. The bottom line of any BCP is the relationship between information access and revenue. A successful BCP should consider and proactively address all possible disruptions to an organization`s information flow. To do this, the plan must include an enterprise storage solution with the following characteristics:

- A remote mirrored data facility that is synchronized to "live" production data: Mirroring data is not a new concept. In fact, it has been around for almost a decade. If a drive or an array fails, a protected second copy of the data exists. This concept has been taken a step further: The mirrored copy is now located at a remote data facility--in the next room or across the country. In the event of a planned (e.g., daily back up) or unplanned disruption of information flow, enterprise storage should enable a fast switchover to the remote copy of the production site.

- Creation and use of multiple real-time copies of production data: When mirroring was first introduced, IT managers balked at the expense of the process. Today, managers sometimes demand not one but multiple copies of their production data to back up critical data, refresh data warehouses, test new applications, implement decision support, achieve Year 2000 compliance, or deal with the European currency conversion.

- By creating additional real-time copies of their production data, IT managers can run production applications and point-in-time backups simultaneously. Armed with multiple copies of their data, managers can schedule multiple processes against the same data in parallel. This effectively eliminates the backup window and enables 7x24 availability for critical databases.

- The ability to migrate data between different hosts and operating systems: Support of possible data center expansion and system upgrades is an important component of a BCP. Businesses must continue through changes and advances in technology as well as through IT shake-ups resulting from corporate mergers and acquisitions. Companies cannot afford to disrupt their operations to migrate information from older mainframes to newer Unix or Windows NT environments. Specialized enterprise storage software allows data to be moved from one storage system to another without affecting information access.

- Automatic failover and load-balancing capabilities: A BCP relies on software that automatically and efficiently balances workloads between the server and the enterprise storage system while offering path failover to alternative data paths. These qualities ensure access to mission-critical applications and uninterrupted business operation.

- A simple method of testing the BCP without disrupting the information/revenue stream: The BCP must have a built-in test and audit process. Full testing is rarely done today because of the risk involved with "pulling the plug" or the cost of establishing a separate test facility. Enterprise storage and remote data mirroring software make data replication easy, allowing full production between two sites. Testing amounts to simply "throwing the switch" and moving primary operations to the other location. Switching back can be accomplished in a matter of minutes.

To be successful, these software and service attributes must be supported by an enterprise storage hardware solution that meets the industry`s highest standards for system and data availability. Ideally, business continuity and high availability should be a primary objective, using RAID protection schemes and full component redundancy.

An effective enterprise storage solution must also be supported by around-the-clock customer service and have the internal ability to self-diagnose to avoid potential disruptions. Background diagnostics, which periodically test and analyze system performance (without degrading performance), reduce vulnerable periods. Business continuity, and ultimately revenue continuity, depends on such a comprehensive enterprise storage strategy.

Click here to enlarge image

According to a survey conducted by Find/SVP, IT executives believe enterprise storage offers solutions to achieve improved availability.

Click here to enlarge image

Regarding information protection, IT executives are most concerned about production system data availability.

Taking Stock in Business Continuity Planning

Not long ago, financial institutions like the Philadelphia Stock Exchange (PHLX) had to live with the fact that even seemingly benign amounts of computing operation downtime, whether planned or unplanned, could cost them their businesses. That`s no longer the case.

Each day, securities worth millions of dollars trade hands on the PHLX. The loss of these records for a day, or even a few hours, would seriously cripple the PHLX, its members, and its customers.

Faced with these perils, the PHLX set out to establish a plan that would enable them to restore computing operations within two hours of a business interruption for two former PHLX subsidiaries: The Stock Clearing Corporation of Philadelphia and The Philadelphia Depository Trust Company.

Early on, the PHLX knew a time-consuming tape solution was financially and operationally unacceptable. The PHLX partnered with SunGard Recovery Services Inc., in Philadelphia, to implement a business continuity plan for an extended remote data-mirroring solution.

According to James Grogan, SunGard`s vice president of technology support, "The nature of business continuity and disaster recovery is constantly changing. Back in the late 1980s and early 1990s, disaster recovery was based on hardware recovery. Today, customers require continuous information availability, which translates into zero downtime and instantaneous access to information."

Up and running now for 18 months, the PHLX plan involves an EMC Enterprise Storage remote mirroring solution, engineered in collaboration with SunGard, that virtually eliminates downtime and ensures revenue continuity. EMC Symmetrix Remote Data Facility (SRDF) software copies every mainframe transaction to a duplicate Symmetrix system several miles away at SunGard. By replicating production data at the SunGard recovery facility via high-speed fiber-optic circuits, the PHLX is able to switch over to the remote copy of production data at the SunGard site and begin processing almost immediately, should the system go down at the PHLX location.

According to Joe Mitchell, vice president of technology for the PHLX, "Mirroring production data at a remote location makes it possible for us to have access to mission-critical information from the exact point-of-failure." In other words, PHLX has achieved what they set out to do--they are able to regain access to all production information and be up and running again in as little as 15 minutes from the time of system shutdown at the PHLX. SunGard`s MegaCenters are interconnected with a high-speed SONET ring, using T3 lines in a closed loop with dual paths to the SunGard site.

Also, because SRDF resides within the Symmetrix system, there is no need to dedicate expensive mainframe cycles to the PHLX`s business continuity solution. This frees up the mainframe to focus on running the business.

For the PHLX, a cohesive business continuity plan has had benefits beyond standard disaster recovery. As part of the contract with SunGard, the PHLX is able to conduct various scheduled tests on the system and has decided to use these sessions to conduct Year 2000 testing. To date, running on IBM OS/390, the PHLX has been able to reset the system date for the millennium changeover on the production data at SunGard, and the systems have run smoothly without hesitation.

With their finger-crossing days behind them, the PHLX is prepared to sidestep any interruption to ongoing operations. By implementing a remote mirroring solution based on enterprise storage, PHLX now has an insurance policy for providing the best possible service to its customers through uninterrupted information access.

Click here to enlarge image

Ron Koshko is business continuity solutions manager at EMC Corp., in Hopkinton, MA.


Comment and Contribute
(Maximum characters: 1200). You have
characters left.