Today, IT systems are required to be operational 24x7x365. That’s 8,760 hours, or 525,600 minutes, or 31,536,000 seconds a year. And, they have to perform flawlessly – while based in different geographic locations. And yet, the reality is that as a result of natural disaster, system error, or other factors, all systems will fail at some point. The key to limiting the impact of system failure is to fix the problem fast by providing for rapid (sub-second) failover to alternative data center resources. How organizations achieve that is the challenge. This report provides IT managers with an overview of the requirements for continuous availability and the critical role of the network to provide system resiliency.
Business trends, like increased use of cloud-based applications, business process automation, digital storage, and the globalization of business, have increased the requirements for IT systems. Application downtime is now unacceptable given customer expectations, lost employee productivity and the impact on business partners. Additionally, the needs for disaster isolation, high insurance premiums, and regulatory requirements are pushing IT organizations to distribute their data centers to geographically diverse locations. The end result is that leading IT organizations (including three financial services firms profiled in this study) are implementing continuous availability via migration to an active-active data center architecture. This continuous availability architecture offers IT the ability to support applications independent of data center location.
The continuous availability data center architecture relies on a high-performance, highly reliable wide area network (WAN). Advances in network technologies, including software defined networks (SDN), now enable the network to support continuous availability via geographically distributed data centers. Important factors to consider when architecting the WAN to support continuous availability include:
- High performance and low latency. The network must support high data throughput with very low latency.
- Ability to support fast failover and VM mobility to a geographically distributed data center(s).
- Automated network provisioning, application visibility, and support for service level agreements (SLAs).
Feature image courtesy Flickr user IntelFreePress