Scaling for a Pandemic: How We Keep Ahead of Demand for Google Meet during COVID-19
summary
February–March 2020 incident response organizes 100+ engineers across 10 teams into parallel work streams (capacity, bottlenecks, dependencies, production, comms) to target up to 100x baseline load; actions include modeling sessions to procure compute, finding code hotspots (30% CPU spent exporting metrics), deploying control knobs (lower default stream quality, disable captions) and automating turn‑up steps. Metrics called vitals drive weekly ramp‑down. Result: no user‑facing outage, ~100M daily participants by April and 235M by October.
AI-written summary · about
chapters (4)
related talks
51mPanel: Learning from Adaptations to Coronavirus
32mHow We Survived (and Thrived) During The Pandemic and Helped Millions...
28mTop 5 Things You Can Do To Reduce Operational Load
27mSublinear Scaling in Practice: The 1k SRE Project
39mOops. I broke the Google, now what?
43mGoogle - Load-shedding
16mThe Pandemic and The Classroom—Enabling Education for Millions
46mManaging a Billion Transactions with Tech, Tactics, and Teamwork