Chapter 3. Principle 2: Define Your Workload’s Required Availability
One of the most complex and risky tasks in a datacenter-based application environment is ensuring high availability. Availability could be threatened by fleeting issues, such as network outages or server crashes, or by more long-term issues, such as a datacenter disaster.
Organizations use a wide array of technologies and strategies to guard against the intermittent or long-term unavailability of their systems. Setting up these high-availability systems (such as Oracle Data Guard and Real Application Clusters for databases) isn’t a trivial concern, and it can be hard to configure and maintain a bulletproof system.
A big advantage offered by cloud environments is the ease with which you can ensure high availability for your cloud-based computing infrastructure and services. Since all leading cloud providers employ the concept of geographically isolated Availability Zones and regions to run their computing resources, there’s built-in capability to guard against disasters and against intermittent hardware and software failures.
Provisioning VMs and running workloads on them shouldn’t be your first step when planning for the cloud. Cloud providers use various strategies to enhance the availability and reliability of VMs you run in the cloud. You must plan your VPNs, regions and Availability Zones, load balancers to handle traffic, and many other considerations before you can start working in the cloud. Although some ...