Achieving High Availability with Failover Software
This article examines how high availability can be accomplished using failover software, which Aberdeen Group calls “the one core technology that serves as the foundation for any high availability strategy.”
In a recent IDC survey, IT and business professionals report that “decreasing recovery time” for email, Web and database applications is the top priority over the next 12 months.
One approach for drastically reducing recovery time is with a sophisticated high availability strategy. Among the most ardent adopters of high availability solutions are financial services companies and companies with a highly mobile professional workforce, as found in the legal and medical sectors.
A well-developed high availability plan is mission critical for many companies in order to protect against losses due to failure. Some companies describe their high availability plan as an insurance policy for the enterprise. Without it, an enterprise is exposed to lost sales opportunities, revenue, productivity of mobile workers and executives and, ultimately, damaged brand reputation. With so much at stake, expectation to recover quickly from interrupted service is greater than ever.
Most companies employ at least backup software as a baseline pre-emptive practice to avert data loss during a disaster. However, while suitable for archiving, backup software alone does not have the fast data recovery ability that companies now require for continuous uptime, according to the Aberdeen Group report, “The Importance of High Availability” (March, 2007)
Reliance on replication software alone is also an insufficient method of ensuring high availability.
According to Aberdeen, “The one core technology that serves as the foundation for any high availability strategy is failover software. Installing redundant physical systems throughout an organization is pointless unless there is failover software that acts like a switch; enabling critical applications to continue running and handling I/O workload even when a primary system comes down.”
Failover solutions available today provide mail server and database protection across the entire mobile messaging environment, and reduce recovery time to a matter of minutes. Neverfail’s high availability solution, for example, requires no manual restart of applications by users. This often results in failures going unnoticed by users.
Three Important Steps to Success for High Availability
1. Make high availability a C-level executive priority.
Many initiatives suffer because they are perceived as IT issues. High availability needs to be a top-down objective to ensure the necessary budget and resources are focused in the right places.
2. Invest in failover technology.
Research found that 63% of Best in Class companies invest in failover technologies compared with 44% of Laggards, which are satisfied with snapshot software. While the latter is a suitable technology for some applications, it’s not considered a true high availability technology.
3. Implement a single group with broad skills to oversee and centralize management of the company’s HA/DR/BC initiative.
A unified and centralized approach strengthens a company’s ability to withstand and recover from a disaster or system crisis, while reducing the potential for inefficiencies and overlap.
Source: The Importance of High Availability, Aberdeen Group, March 2007
Top 10 Questions To Ask When Evaluating a High Availability Solution
Will the solution…
1. perform a failover that is transparent to the user — meaning users do not have to reboot their workstation or restart their application?
2. check the overall health of the server prior to installation to ensure ongoing server reliability?
3. feature an installation process that does not require the failover server to be manually pre-configured? Does the installation require any downtime for the protected application?
4. offer complete data protection — ensure both the data and the registry can be rolled back?
5. protect from all types of failures, including failures caused by data corruption or data loss
- operating system performance
- network issues
- hardware issues
- application issues, including stalled, hung or non-performing applications
6. work in a WAN environment to protect from complete site failure?
7. offer both a manual and automatic failover/failback by the click of a single button?
8. be easy to maintain — what IT resources are required during upgrades, failovers and switchbacks?
9. protect the primary application (e.g. Exchange) AND the auxiliary applications “out-of-the-box,” such as anti-virus, archiving, back-up, anti-spam, and fax products?
10. allow for comprehensive failover testing — including the application, Active Directory, DNS updates and network connectivity?
Source: Neverfail Website http://www.neverfailgroup.com/resources/faq.aspx
In addition to fast recovery time, IT and C-level executives that have confidence in the availability of its systems, will be more likely to deploy mission-critical mobile applications, knowing that higher adoption rates, productivity gains and ROI are possible when the chances of downtime due to failure are significantly reduced.
This in turn removes barriers to further expansion of mobile deployments and applications to a larger percentage of the workforce, as there is no longer a concern that users will lose connectivity to those business critical functions.
Taking High Availability One Step Further: Continuous Availability
Some IT departments express the desire to not have to ‘recover’ from an outage at all, but instead avoid outages altogether.
The term “high availability” implies that some degree of service outage is inevitable, whether arising from planned or unplanned outages, and that some amount of system recovery will then follow. In contrast, “continuous availability” solutions, like those available from Neverfail, are designed so that there are literally no circumstances that will cause application service to users to be interrupted.
Neverfail’s solutions continuously monitor application service levels and, when problems arise, provide seamless failover and failback without the need to recover or restart any applications.
Continuous Availability in Action
British American Tobacco (BAT) put Neverfail’s solution to the test in April 2007 when the company conducted electrical maintenance work at its Headquarters over a three-day period. Using Neverfail, the company was able to failover its BlackBerry applications to its disaster recovery site in a matter of minutes and subsequently fail these services back with equal ease, with no interruption to service and, hence, no ‘recovery’ required.
“Previously, this type of outage would have been an absolute nightmare to manage,” says Laurence Dale, Technical Architect of BAT’s BlackBerry solution. “It would have had a significant impact on our IT team’s resources and we would have risked a loss in the BlackBerry service. With Neverfail, this potentially complicated situation was cut down to the click of a few buttons.”
For more on continuous availability solutions from Neverfail, visit www.neverfailgroup.com
High Availability and Business Continuity Planning