Quality Assurance and Testing Strategies for Taobao Double 11 Interactive Games
For Taobao’s 2022 Double 11 interactive games, a comprehensive QA approach combined online stress testing, reinforcement‑learning‑driven case generation, and systematic fault injection—addressing high‑concurrency, data‑consistency, and client‑performance risks through traffic shaping, data and middleware isolation, and automated tools (Wali, DaBai) that delivered over 60 smoke cases, 76 % coverage and 82.6 % fault detection.
Background: In 2022 Taobao Double 11, three interactive games (single‑player mowing, team PK, price‑guessing) were launched to meet diverse user needs.
Core strategy: Use online stress testing to adjust service capacity, reinforcement learning to cover test paths, and system fault injection to discover corner cases, ensuring systematic quality assurance.
Key challenges:
High concurrency and strict timing in the price‑guessing game leading to user aborts.
New PK model introduces data consistency risks (serialization protocol, player‑team‑room storage redesign, RDB stability).
Real‑time interaction raises client performance requirements (crash, rendering, latency).
Client‑side logic for single‑player mowing adds new consistency and security concerns.
Risk assessment (selected items):
• Single‑player mowing – energy collection – High risk: functional, collision, fault, compatibility tests; indexDB/localStorage stability; client crash.
• PK model – medium risk: functional and fault tests.
• Serialization protocol – medium risk: size, efficiency, compatibility.
Testing strategy:
Online pressure testing with capacity‑aware traffic shaping and limit‑rate controls.
Data isolation via shadow tables, separate RDB instances, LDB shadow links, and tagged local cache.
Middleware isolation by versioning and group‑ID segregation.
Ensuring test interfaces reflect real‑world upstream/downstream dependencies.
Automation tools:
“Wali” robot – a reinforcement‑learning based engine that automatically generates end‑to‑end PK test cases, achieving 76 % of test cases as automated smoke tests and 100 % pass rate.
“DaBai” platform – automatically creates business‑exception scenarios without scripting, using flow‑analysis, JVM‑Sandbox, and traffic isolation to produce high‑coverage fault cases.
Results:
60+ smoke cases generated, covering 76 % of total cases.
216 automated disaster‑recovery cases; P1/P2 fault coverage 82.6 %.
Future outlook includes expanding automated test coverage, improving productization of the exception platform, and enhancing real‑time game architecture research.
Signed-in readers can open the original source through BestHub's protected redirect.
This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactand we will review it promptly.
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.