6 months ago
2
Thousands of users crossed India recovered themselves abruptly locked retired of Google Meet calls aft what appears to person been a wide-scale work disruption. Reports clustered astir mistake messages and “502 — Bad Gateway” notifications, with galore users incapable to entree meetings.
The outage struck conscionable days aft a monolithic disruption triggered by Cloudflare, underscoring however profoundly galore fashionable online services beryllium connected a fewer infrastructure providers.
While Google has not yet released a elaborate study connected the outage, the timing and mistake patterns rise concerns among galore users arsenic truthful galore platforms trust connected the unreality and content-delivery infrastructure.
The Cloudflare outage connected November 18 had already exposed this vulnerability. On that day, Cloudflare reported that a bug successful its Bot Management strategy caused a configuration record to balloon beyond expected limits.
For galore organisations and millions of extremity users, Cloudflare efficaciously acts arsenic the “front door” of their online presence, providing caching, security, routing, DDoS protection, and planetary distribution.
Just a period earlier the Cloudflare meltdown, AWS, the world’s largest cloud-compute and hosting provider, experienced its ain large disruption originating successful its U.S.-EAST-1 region. A DNS-related responsibility triggered cascading issues crossed thousands of services babelike connected AWS, causing prolonged downtime for apps and sites astir the globe.
When seen together, these incidents overgarment a worrying picture: the precise systems designed to marque the net much resilient, scalable and accessible are progressively becoming captious azygous points of failure.
