At 12:20 PM IST, we started facing major outages across all our processing regions and servers. All uncached requests (missed from CDN and our internal caches) were not getting served during the impacted duration. The issue was partially resolved at 12:40 PM IST and completely resolved at 12:45 IST.
The issue resulted from a configuration change on our servers that got picked up around 12:20 PM IST when the issue started happening. Once our team identified the error, the configuration change causing the outage was handled on our servers as a temporary measure, and the service was restored.
Soon after, our technology team started making changes to ensure that such configuration changes do not impact the server uptime in the future. Our team will soon deploy the change across our systems to prevent this issue from recurring in the future. Meanwhile, the systems are completely operational, and our team is constantly monitoring the same.