How JD Health Engineered System Stability for the 618 Mega‑Sale

Background and Business Context

In 2020, the COVID‑19 pandemic accelerated the growth of online health services, making the 618 shopping festival a critical moment for JD Health. The company needed to handle massive traffic spikes while supporting multiple business lines such as large pharmacies, internet hospitals, drug procurement, and consumer health services.

Technical Challenges

New business operations, system expansions, and continuous demand creation introduced risks to system stability and performance. From late May to mid‑June, a series of high‑traffic, large‑scale system upgrades and migrations threatened both reliability and latency.

Preparation Strategy

The R&D team adopted a multi‑layered preparation plan:

Comprehensive Rehearsal : Conducted full‑scale dry runs to anticipate issues before the promotion.

Stress Testing : Estimated 20‑times normal load for each interface based on historical 618 traffic data.

Architecture Review : Classified 52 systems into three levels (11 level‑0, 13 level‑1, 28 level‑2/3) and performed dependency checks, capacity planning, and optimization.

Risk Mitigation : Developed downgrade plans, introduced chaos‑monkey testing, and performed cross‑team inspections.

Dual‑Channel Risk Control : Daily risk‑assessment meetings, reporting by system owners, and centralized resolution by a dedicated steering committee.

Real‑Time Monitoring : Enforced strict code reviews and impact assessments before deployment, and set up 24‑hour monitoring with rapid incident response.

Execution and Outcomes

All new features and system changes passed architecture committee reviews and were deployed only after meeting stability criteria. Continuous online inspections and three rounds of pre‑sale testing ensured early detection of anomalies. During the 618 event, the team maintained 24‑hour rotating on‑call duties, achieving stable system performance and swift issue resolution.