Downtime sucks. Avoiding downtime for anyone, individual or organization, is about as essential as pizza to a developer. High-availability of cloud services is critical in the current age of cloud-driven everything. Service Level Agreements (SLAs) are often the tie-breaker when selecting best of breed cloud platforms in the market. Vendors experiencing unexpected outages find themselves thrown to the wolves – unsurprising considering the cost of systems downtime averages $7,900 per minute according to one research report.
Fortunately, the answer to availability concerns in the cloud lies within cloud computing itself – in the form of high-availability cloud architecture.
High-availability cloud architecture separates services and applications from the underlying cloud infrastructure, eliminating any single-point-of-failure. The infrastructure itself comprises specific network configurations, redundant infrastructure components, software-defined networking, separation of storage and compute and a multitude of other aspects architected together for continuous availability.
In the event of a node outage, for example, workloads are automatically migrated to find and run on any available components, with performance unaffected. Our ‘self-healing’ architecture means that if our software-defined storage system detects that a node is having an issue or is offline, data replication of the missing data blocks begins immediately and entirely automatically. Our separation of storage and compute also allows a VM to be provisioned on a different node whilst maintaining access to the same storage.
The beauty about infrastructure-as-a-service is that individuals and organisations can access this level of resiliency without spending money on physical infrastructure.
Besides minimal downtime, users also experience faster disaster recovery, the ability to conduct scheduled maintenance with no downtime, fault isolation, live VM cloning and migration and independent scaling of tiers.
So what can the user do to maximize the value of high-availability architecture? Here are a few tips:
And of course, thorough testing of all components and applications is critical.