Redundancy is one of most important ways to improve the reliability and availability of a system. But it’s easy to think you have redundancy when you really don’t. The system has to be configured in such a way to deliver true redundancy.
It’s a lot like Noah’s Ark – you have two of everything throughout the stack and you think you have achieved redundancy. But having two of everything in your system doesn’t mean those two things are working together in such a way where failure and recovery doesn’t cause interruptions. There comes a time when it makes sense to sink Noah’s Ark and build a new boat that is highly efficient and agile. Intelligent Availability® (IA) is this new boat. It’s proactive system management. IA holistically collects, interprets and adapts itself to protect system uptime. One of the key components of IA is full stack redundancy. IA requires complete hardware redundancy from top to bottom. From compute nodes to networking/switches to power input, there are two of everything. Traditional High Availability systems often miss redundancy of power, networking and the management layer. Even Noah – try as he surely did – was not able to achieve true redundancy. He missed a few animals when he was loading up the ark. That’s probably why you don’t see unicorns anymore.
When looking for a blueprint from which to build a new boat (or new server system), look at Intelligent Availability® to be your guide. IA closes all of the gaps such as redundancy and protects server uptime. It’s peace of mind so that if a biblical flood ever swamps your server room, you can recover quickly and predictably with minimal to no downtime.