Product Reliability for Google Maps
summary
Maps reliability shifts from server SLOs to direct user signals for a billion monthly users, run by a ~20‑person SRE team, claiming it catches issues server metrics miss. They build user experience signals (availability, latency, data quality), ingest visual‑element impressions, use traces via peretto and automate change supervision across server rollouts, app binary releases and feature ramps. A defense and depth system prevents the annual “zero results” outages previously escaping detection, though it initially surfaces more incidents and raises debugging costs.
AI-written summary · about
chapters (16)
- 0:00Product reliability at Google Maps
- 1:30Motivation for reliability work
- 3:18Gaps in server alerts
- 4:01User outages examples
- 7:33User-centric reliability focus
- 8:34UX monitoring strategy
- 9:39Engaging product teams
- 10:43Direct measurement implementations
- 12:05UX availability measurement
- 13:30Latency tracking
- 14:54Data quality monitoring
- 15:57Change supervision
- 17:44Case study: reliability improvements
- 19:50Lessons from UX monitoring
- 23:45Future work: root cause analysis
- 24:49Q&A
related talks
38mSRE for Mobile Applications
40mHow Google SRE and developers work together
35mSRE 101
40mWhy Is My App SLOw? Defining Reliability in Platform Engineering
32mHow Google SRE and Developers Work Together
18mProduct Reliability in
14mWeb Apps, through the Lens of SRE (Site Reliability Engineering)
41mAn Introduction to SRE at Google (with DevOps Focus)