Design and Implementation of a Cluster‑Aware Guava Cache Component for High Reliability
The paper presents a cluster‑aware Guava cache component for Alibaba’s Xianyu platform that mitigates downstream service failures by adding asynchronous reload, cluster‑wide key invalidation, and size reporting, enabling automatic fallback to refreshed local data and improving latency, with future plans for a management console, tiered storage, and disk‑backed caching.
