Key Ops and Cloud Takeaways from Our Tech Exchanges with AutoHome & Sina

Background

In the spirit of open collaboration, the addops team organized technical exchange sessions with AutoHome and Sina, focusing on practical, in‑depth discussions rather than high‑level conference talks.

AutoHome Session

We invited AutoHome’s operations team. Our “Zhe Ye” introduced the Hulk cloud platform, highlighting CMDB design for handling complex business lines.

AutoHome shared their CMDB system with visualized rack and network topology, and demonstrated their code release system capable of handling thousands of nodes.

We presented our containerization experience, building a highly available Kubernetes‑based architecture for diverse user scenarios.

AutoHome also discussed their use of Radar for Redis cluster management.

Our team showcased the “wonder” monitoring system built on open‑falcon, supporting tens of thousands of nodes with flexible alerting and monitoring policies.

AutoHome then shared their own monitoring system, including a directed acyclic graph analysis for complex service dependencies.

Sina Visit

After the AutoHome session, Sina invited us for a technical exchange. We toured their new headquarters and shared load‑balancing management and architecture experiences for layers 4/7.

We discussed containerization practices, and Sina presented their Docker‑based hybrid‑cloud architecture and the DCP system that addresses deployment and monitoring challenges in hybrid environments.

We also covered virtualization applications in business, and shared our experiences with machine‑learning techniques applied to operations.

Conclusion

The addops team, through the Hulk cloud platform, continues to improve operations services and quality, and welcomes further technical exchanges with peers.