Replatforming Online Search for a Major High Street Retailer
summary
A product team replaces a proprietary on‑prem search with cloud microservices using Elasticsearch and Google Cloud, deploys infrastructure as code with Terraform, and goes live for Black Friday 2019. They start with an A/B cut of one percent to 100% traffic (one percent in July, full 100% by mid‑2019), slash a 90‑minute legacy reindex to seconds, operate daily deploys, run over 100 A/B tests, and decommission the old system with improved latency and availability while building in run‑booked 24/7 support.
AI-written summary · about
chapters (19)
- 0:00Intro
- 0:48Story
- 1:19John Lewis
- 3:51Peak Trading
- 5:41Subplots
- 6:13Map
- 6:49Discovery
- 7:20Managed Services
- 11:10The Team
- 17:30The Search Service
- 24:00Keeping Things Simple
- 24:35Live Before Peak
- 29:32Support Through Peak
- 31:24Building on Peak
- 35:46Change Management
- 40:13Build It Run
- 42:23Dashboard
- 45:43Remote working
- 47:23Conclusion
related talks
38mScaling an online search engine to thousands of physical stores
35mInnovating at Scale: Lessons from Pegasystems' OpenSearch Journey
42mBuilding a real-time news search engine #bbuzz
22mEvolving Search Architecture for a Demanding Tesco Online Business
28mNavigating the Transition: SRE Challenges and Highlights
28mEnhancing Elasticsearch Performance: Innovative Reindexing...
01h 07mMicroservices from the Trenches. A Guide through the Architecture, Design, Code
31mHow technology driven B2B disruption works