It's a well worn fact. For IS executives, disaster recovery means the restoration of computing operations after a natural or manmade disaster, be it an earthquake, a tornado, a terrorist attack, or a power outage. In the past, companies had no other option than to 'recover' from disasters. They were forced to accept not only the inconvenience of disruptions, but also the costs. Today, the notion of disaster recovery assumes an unacceptable level of business disruption was allowed to occur following a system interruption. Both the internal and external customers served by IS expect uninterrupted world class service and IS is expected to deliver.
The focus of individuals on both the technical and business side of operations has evolved from disaster recovery to the more contemporary business continuance model. Business continuance deals not with recovery, but instead on the ability to manage and ensure the constant availability of information across the computing enterprise. Business Continuance ensures that events caused by natural disasters, or non-productive interruptions, occur behind the scenes, while bottom-line, revenue generating applications are not interrupted.
- A cohesive business continuance strategy concerns not only unplanned disasters, but also planned interruptions. Access to production data must also continue through routine and planned events such as point-in-time backups, data warehouse loads, data migration, data transfer and system testing for timely dilemmas like year 2000 testing and eurocurrency conversion.
A recent 1997 survey of 700 IT executives worldwide, conducted by Find/SVP, revealed that information system downtime (planned or unplanned) continues to be a widespread issue throughout data centers today. Sixty one percent of the IT executives polled expressed the need for the fastest possible solution for disaster recovery. In addition, 43 percent of the respondents reported more than five incidents per year, while 13 percent reported that they experienced unplanned downtime that lasted more than a day. These executives seek a business continuance solution to eliminate downtime and protect critical information from any business disruption.
Enterprise Storage: Creating A Business Continuance Infrastructure
Enterprise storage has emerged as the key element in a successful business continuance infrastructure. It provides a common environment to protect against system downtime, while ensuring bullet-proof business continuance protection. Enterprise storage consolidates information to give IS a common view of their data. Enterprise storage enables them to protect, manage, share and make information constantly available to users. While storage was once regarded as a device merely used to archive information, it is now a recognized pillar of today's IT infrastructure, bridging the gap between incompatible computing platforms and allowing the sharing of business information across the enterprise regardless of platform.
Twice is Nice: Using Information Mirroring Technologies
Any IT executive will agree that two copies of data are better than one. However, the perception that mirrored copies will be twice as expensive is simply not true. Implementation of disaster recovery/business continuance solutions is best achieved with the combination of highly specialized enterprise storage hardware and software. In the event of a data center disruption, traditional methods of restoring data involved replicating the data to hundreds and thousands of tapes. These copies were then either electronically or physically transported to a remote site for storage. Time consuming restoration of the data and applications logs made this a fairly cumbersome method, and as result, is practically obsolete.
In contrast, a comprehensive business continuance solution will combine enterprise storage hardware and software technologies to achieve superior levels of information availability and accessibility during planned or unplanned outages. A viable business continuance solution must implement mirroring technologies to provide an online, host-independent method for duplicating production site information over long distances. Mirroring techniques 'replicate' or migrate information from one 'Source' storage system to another 'Target' storage device in a geographically remote location. By providing full mirroring of production data between the systems, companies can immediately switch over to the second copy of data and be up and running again in minutes - with little or no downtime.
Testing Disaster Recovery
By mirroring data at remote locations, business continuance plans can be tested without intense manual recovery drills, complex procedures, and/or business interruptions. The IS staff can switch computing operations to a remote site at any time, and testing can be conducted during normal business hours. This is especially critical for businesses serving global markets that operate 24 hours a day, seven days a week.
By mirroring all user specified information from a 'Source' storage device to the 'Target' storage system, IS staff can quickly resume operations on the Target system when data center operations are disrupted. If an outage occurs at the central storage Source site, an I/O operation can be completed from the Target storage system. When the communication link is re-established, there is an automatic resynchronization of data between the storage systems and normal operations can resume. Today this 'campus-like' solution can be achieved remotely (10 to 66 kilometers from the central data center) as connectivity is provided across standard mainframe ESCON fibre and open system connections. For companies that require remote sites farther away geographically, extended distances must be supported across broadband T3 & E3 telecommunications links.
Twice is Nice, but Three or More would be Ideal
The increased adoption of 24x7 business operations' especially for global enterprises or government agencies - leaves little time for important tasks such as nightly database backups or Year 2000 testing. Another key technique to further extend disaster recovery/business continuance is the use of 'multiple mirroring' software technology that enables IS executives to non-disruptively create multiple copes of mainframe or open systems production data. As previously discussed, traditional data mirroring involves creating a mirror image of data on a second disk. The multiple mirroring capability extends traditional mirroring techniques by creating independently addressable copies of production data, or Business Continuance Volumes (BCVs).
BCVs are created within the storage system and backup operations can be performed against it, leaving the production data available for regular business processing. Applications that might otherwise shut down, or slow down, can run continuously, enhancing a corporation's ability to earn revenue and serve customers.
BCVs, or copies of real production data, can be used to run simultaneous tasks in parallel. This parallel processing capability is known as 'workload compression.' Once the BCVs have been created, they can be further split from the mirrored volumes for tasks such as loading data warehouses, testing new applications, Year 2000 testing, running batch jobs, daily backups and any functions that typically disrupt ongoing, normal business processing.
By combining traditional mirroring technologies with multiple mirroring technology (BCVs), companies can achieve the highest-level security. In the event of an unplanned interruption, downtime can be reduced or virtually eliminated and more importantly normal computing operations can be immediately resumed.
Jim Ogorchock, Business Continuance Manager at EMC Corporation, is responsible for marketing strategy, marketing programs, and strategic alliances related to the company's business continuance efforts.- EMC Corporation is based in Hopkinton, Massachusetts, and is the world's leading supplier of intelligent enterprise storage systems and software for mainframe and open systems environments. For further information about EMC products and services, EMC's home page can be accessed at http://www.emc.com.