The article describes how Hujiang built and iterated a three‑generation asynchronous task scheduling platform—Juice—using Mesos and Docker to achieve on‑demand resource allocation, horizontal scalability, and efficient handling of large multimedia processing workloads.
This article details Dangdang’s experience building a private job‑cloud platform using Mesos and a custom framework, covering technology selection, key features, integration challenges with Elastic‑Job, resource scheduling, and practical recommendations for Mesos‑based Java workloads.
This article explains the Mesos distributed system kernel, its resource‑allocation workflow, and how Qunar engineers applied and evolved Mesos, Marathon, and custom frameworks to achieve fine‑grained scheduling, high availability, service discovery, and multi‑tenant management in a large‑scale production environment.
This article examines resource allocation philosophies—auction, budgeting, and preemption—and compares the architectures, data models, and APIs of major schedulers such as Borg, Omega, Mesos, Kubernetes, and Alibaba’s Zeus, while also exploring sharing strategies, task classifications, utilization metrics, and predictive techniques for efficient resource management.
This article explains how large‑scale elastic architecture designs achieve high availability and scalability by automatically allocating resources during traffic spikes such as Spring Festival ticket sales and Double‑11 shopping festivals, and outlines key techniques, trends, and disaster‑recovery strategies.
This article provides a comprehensive overview of Google’s Borg system, detailing its purpose, user perspective, workload types, cluster and cell architecture, job and task management, scheduling algorithms, scalability techniques, and availability mechanisms for large‑scale distributed environments.
This article explains what a deadlock is, why it occurs, the four necessary conditions that enable it, and outlines common strategies—including ignoring, detection and recovery, avoidance, and prevention—to manage or eliminate deadlocks in operating systems.