Preparing for a Crash in the Cloud
It seems as if there's no information that can't be moved to the cloud. But what happens when the cloud goes down, as it recently did for Amazon? More than a few IT managers consider this their worst nightmare.
Depending heavily on the cloud is a mistake if you don't have a backup plan in place.
Stan Klimoff, director of cloud services for GridDynamics, gives IT TechNewsDaily readers a few tips on how to prepare for the eventuality of a crash in the cloud. GridDynamics is a tech consultancy that uses the latest advancements in grid and cloud technology to help companies deliver business systems that handle peak loads, scale on demand, and always stay up.
Design for failure. Follow the rule that everything that can fail will fail. This will help your team better prepare for possible outages. Designing for failure is an expensive engineering practice, but it definitely pays off in the end.
Institute disaster drills.Drills provide a good understanding of the weak points of the system, as well as prepares the staff to face a real disaster.
Have a "Plan B."You have to be able to take your stuff and move away from an ISP within hours. Design your system to avoid vendor lock-ins.
Know your Service Level Agreements.The recent cloud outage that Amazon experienced was actually caused by the EBS (Elastic Block Store) service, which has the stated availability on par with a local disk. No one should trust a single local disk with a critical database.