1 views collected around this technical thread.
This article presents a comprehensive guide on applying Linux Cgroups to isolate MySQL resources, covering CPU, memory, I/O, and network bandwidth controls, detailed configuration steps, performance testing with sysbench, and practical observations to ensure stable and predictable database behavior.
The article explains fault isolation as a key technique for improving distributed system availability, detailing multiple isolation levels—from data‑center to user‑level—and complementary strategies such as circuit breakers, timeouts, fast‑fail, load balancing, caching, and degradation switches.
The article explains the mGPU solution from Volcano Engine, detailing its kernel‑level GPU virtualization, container runtime hooks, and scheduling mechanisms that allow multiple containers to share a single NVIDIA GPU with isolated compute and memory resources, achieving near‑lossless performance and up to 50% higher utilization for AI tasks.
This article explains the mGPU solution that virtualizes NVIDIA GPUs for containers, detailing its driver architecture, compute and memory isolation mechanisms, performance benchmarks on ResNet‑50 inference, and how it boosts GPU utilization by over 50% for AI and high‑performance computing tasks.
The article analyzes why deploying MySQL in Docker containers can lead to data‑security, performance, state‑management and resource‑isolation problems, while also outlining scenarios where containerizing MySQL might be feasible and offering practical mitigation strategies.
Vivo’s online‑offline co‑location platform consolidates latency‑sensitive online services and batch offline workloads on shared Kubernetes nodes, using differentiated resource views, priority‑based QoS, and safety watermarks to boost CPU utilization from 13 % to 25 %, adding 20 000 cores and 50 TB memory for peak‑hour offline tasks.
The article examines the challenges of running MySQL in Docker containers, highlighting data‑security, performance, state, and resource‑isolation concerns, while also presenting scenarios and strategies where containerizing MySQL can be viable.
This guide explains Kubernetes namespaces, their purpose for multi‑tenant resource isolation, describes the four default namespaces, and provides step‑by‑step commands to list, create, format, and delete namespaces using kubectl.
This article presents a comprehensive case study of WingPay's big data platform, covering company background, data development and governance platform design, task development workflow, architectural choices, scheduling engine selection, data bus implementation, resource isolation, quality monitoring, cloud‑native practices, future challenges, and a Q&A session.
This article explains how ByteHouse, an enhanced version of ClickHouse used at ByteDance, adds full upsert support, multi‑table joins, high‑availability features, and, most importantly, a Resource Group mechanism that provides fine‑grained CPU, memory, and concurrency isolation to improve query performance and stability.
This article details the design, evolution, and core capabilities of 58.com's self‑developed Star River data scheduling platform, covering its positioning, architectural challenges, high‑availability master design, intelligent monitoring, baseline management, and future roadmap for big‑data operations.
This article explains how ByteHouse extends ClickHouse with a Resource Group mechanism that provides fine‑grained concurrency, memory, and CPU isolation, improving query latency, reducing variance, and increasing cluster stability for large‑scale ad‑tech workloads.
This article explains why resource isolation is essential in distributed systems, outlines various isolation methods such as thread, process, cluster, and data read/write isolation, and demonstrates practical implementations using Netty, Dubbo, Tomcat, and Hystrix with code examples and configuration tips.
This article explains how Linux namespaces achieve resource isolation and how cgroups enforce resource limits, covering the six namespace types, the system calls used to create them, and practical examples of CPU share configuration with Go code and command‑line tools.
The article analyzes why running MySQL in Docker containers can cause data loss, performance bottlenecks, state management problems, and insufficient resource isolation, while also offering scenarios and strategies where containerizing MySQL might still be feasible.
This article explains how Linux namespaces provide resource isolation while cgroups enforce resource limits, detailing their types, system calls, configuration steps, and practical Go code examples to demonstrate CPU share control in container-like environments.
The article examines why containerizing MySQL with Docker is problematic, highlighting risks to data safety, performance bottlenecks, state management challenges, and insufficient resource isolation, while also noting scenarios where containerized databases may be viable.
This article explains the role of Linux cgroup and namespace technologies in providing resource isolation and security for containers, traces their historical development from early chroot mechanisms to modern Docker and Kubernetes, and details cgroup architecture, core files, migration, delegation, and practical usage examples.
This article demonstrates how to achieve multi‑tenant resource isolation in ClickHouse by configuring XML files, leveraging Quotas and Settings Profiles, and transitioning to SQL‑driven access control with roles, illustrating practical steps, code snippets, and best practices for secure user management.
The article explains the importance of isolating resources such as CPU, network, and disk in distributed architectures, describes thread, process, cluster, data‑read/write, static, and crawler isolation methods, and provides concrete code examples and best‑practice recommendations for backend developers.