It used to be reasonable and achievable to recover from a disaster in a day or two. As the ferocity of business competition and pace of business activity has soared, however, the window for disaster recovery has greatly shrunk. Powered largely by evolving technologies, disaster recovery has given way to business continuity as a top priority for IT staff. After all, in a global, 24/7/365 business environment, it doesn’t really matter how slick or advanced your systems and applications are if employees, customers and business partners can’t access them.
The ability to limit system downtime to an imperceptible blip is increasingly becoming IT’s most mission-critical responsibility. The consequences of even a few minutes of system downtime can result in permanent data loss! This is not an overstatement. Gartner estimates that the average company experiences 87 hours of downtime per year at a cost of $42,000 per hour. For companies that are exceptionally technology driven, in which every millisecond counts, downtime costs can easily escalate into the range of millions of dollars per hour.
Beyond the immediate dollar losses, companies that cannot ensure continuous application and data availability stand to lose customers, brand equity and additional costs in the form of fines and other regulatory penalties.
Even with the possibility for such staggering losses, it is astonishing to learn that although most companies now have formal data backup plans in place, those plans are infrequently updated and tested. In the event of a serious IT outage, an alarming number of businesses would grind to an abrupt halt.
Let’s assume you’re not like those companies. You’ve taken the continuous application availability imperative seriously and have implemented (probably at significant expense) a business continuity system designed to ensure zero downtime and zero data loss. But are you ready to bet your career that it will perform as intended, should the need arise. And make no mistake; sooner or later, your system will be tested. Every enterprise is vulnerable!
Conventional Either/Or Solutions
Typically, to ensure continuous application availability, organizations have relied on one of two approaches: storage replication or database replication.
To achieve the continuous availability, many organizations have deployed storage replication systems. Also referred to as disk mirroring, this approach requires the deployment of fully redundant failover sites configured with identical hardware and software environments. Such mirrored systems can be synchronous or asynchronous. While they protect the entire storage system including data, software, and operating systems, they are not completely foolproof and have a number of inherent downsides:
- They are expensive, as the mirrored systems must be identical to the primary systems, and cannot be used for other purposes such as reporting and decision support when they are not in failover mode
- In mirroring an entire environment, disk corruption can also be mirrored, resulting in a perpetuation of the data corruption
- Recovery time can take hours
- Mandated audit trails can be compromised
- Network traffic can be quite intensive and costly
Alternatively, other organizations employ database (also called transactional) replication technology to address their mission critical business and regulatory requirements. This methodology works by reading primary transaction logs, transforming the logs into SQL changes and applying those changes to remote standby databases. Database replication is effective, protecting against disk corruption while ensuring that the standby databases are always available. Additionally, the replicated environment needn’t be identical to the primary environment, allowing organizations to create less costly business continuity systems. When not in failover mode, this approach allows standby databases to be used for other purposes such as decision support and reporting, thereby enhancing organizational productivity and providing a faster ROI on the business continuity investment.
Of course, database replication, too, has drawbacks. The biggest of these is that it employs asynchronous replication, meaning there is the possibility of data loss in the event of a primary systems outage.
The Best of Both Technologies
The optimal approach, given the pros and cons of both storage replication and database replication, would be to implement a hybrid solution, combining the capabilities and benefits of both technologies.
- Protect your entire data infrastructure
- Operate synchronously to protect against data loss
- Minimize network traffic by mirroring only the database transaction log
- Protect against disk corruption
- Maintain transactional integrity
- Enable near instantaneous failover
- Allow the use of less expensive hardware and software for the failover environment
- Allow the failover database to be used for reporting, decision support and other tasks when not in business continuity mode
- Provide architectural flexibility through support for mission-critical databases, synchronous and asynchronous replication, and active/active system configurations
For companies requiring continuous availability of its systems and access to terabytes of data collected and maintained in those systems, a hybrid solution is ideal. Implementing a storage-based business continuity solution at a remote location can ensure the continuous availability of a system as well as the safety and integrity of up to terabytes of data.
Companies with very demanding requirements can still keep their systems up and running by combining the best attributes and capabilities of storage and transactional replication. Capabilities include resuming operations rapidly, preventing data loss and ensuring complete data integrity.
By developing and deploying a hybrid business continuity and disaster recovery solution, failover time is minimized; guaranteeing zero data loss and ensuring complete data integrity. Businesses can also use a failover environment under normal conditions to run reports and perform other database tasks, generating a return on their business continuity investment.
Let’s paint another scenario to address many companies’ concerns about keeping operational systems up and running, and protecting critical business information.
When dealing with situations where application downtime can cost great financial loss, a company needs a system that can guarantee a complete recovery and enable continuous system availability and data protection. A common issue, especially when dealing within the financial markets industry, is the need specifically for the ability to squeeze additional value out of a failover environment by allowing those assets to be used for reporting and decision support activities when not being used for business continuity. Addressing application availability and zero data loss requirements with a hybrid approach establishes a failover solution with zero data loss and can complement an existing storage replication system.
Businesses can see dramatic reductions in failover time and restored availability in seconds with no data loss. It is also possible to reduced network bandwidth costs and make use of the hardware and software that normally lie dormant when the primary systems are up and running to run reports and do analysis, rather than have them just gathering dust while waiting for a failure to occur.
By implementing a “best of both worlds” solution is great for companies needing to address disaster recovery and back-up needs.
Kiss Technology Trade-Offs Goodbye
Most people understand the necessity for solutions that ensure zero downtime and zero data loss in the financial services world. But those requirements are growing equally pervasive and critical across a spectrum of industries including healthcare, retail, telecommunications, manufacturing, energy, transportation and others. It’s simply a consequence of an IT driven and dependent economy.
No one understands the pressure to implement such solutions better than you and your colleagues. However, do they really understand the technical considerations and the consequent trade-offs you’ve been forced to make in addressing your organizations’ business continuity requirements?
You can kiss those trade-offs goodbye. By complementing your existing storage replication solutions with the additional capabilities and benefits inherent in database replication you can:
- Ensure failover within seconds so your end-users are unaware of the problem,
- Lower your total cost of ownership by reducing network bandwidth requirements and minimizing hardware costs at your failover site,
- Increase your ROI by making your standby database available for reporting and decision support (rather than having it sit idle waiting for a failure to occur)
- Feel confident in saying, “Our systems are up and running 24/7/365.”
Bill Zhang is the senior product manager of replication solutions for Sybase, a leader in the industry in working to ensure business continuity.