The current Google Cloud incident in London is the sort of occasion that ought to pressure enterprises to rethink what cloud resilience actually means. Google stated the disruption was attributable to a “energy occasion” affecting London-based knowledge facilities in its europe-west2 area. The difficulty appeared early within the day and affected cloud companies that many organizations depend upon, both immediately by means of their very own deployments or not directly by means of third-party platforms that run on Google Cloud.
What made this incident vital was not merely {that a} cloud area had issues. Cloud infrastructure, like all infrastructure, can and can fail. The actual situation was the scope of the dependency chain. Banks, supply companies, e-bike suppliers, budgeting apps, some appointment companies by means of the Nationwide Well being Service app, and courtroom video-link programs had been all caught within the disruption. These companies could seem unrelated, however in actuality, they’ll share the identical underlying cloud suppliers, regional infrastructure, managed companies, networks, id programs, APIs, and operational dependencies.
Google later stated companies had recovered for many affected workloads, with solely a small variety of prospects nonetheless impacted. Nonetheless, the harm had already been performed. The incident confirmed how a lot trendy enterprise is dependent upon cloud platforms which might be extremely succesful however not resistant to failure. It additionally confirmed that resilience will not be one thing enterprises mechanically achieve by shifting to a hyperscaler. Resilient structure should be designed, funded, examined, and owned.