0 views collected around this technical thread.
This article explains the network interruption issues of VM live migration in 360's VPC overlay (V1), analyzes the root causes, and presents a V2 redesign that eliminates downtime through pre‑loaded forwarding policies, traffic redirection, and streamlined component collaboration.
This article explains how to achieve seamless code updates for SpringBoot applications by allowing two instances to share the same port, detailing the underlying Tomcat embedding mechanism, DispatcherServlet handling, and providing a complete Java implementation with step‑by‑step instructions and test results.
This article explains how to enable two SpringBoot processes to share the same port for seamless code updates, covering the underlying servlet container mechanics, the required initialization flow, and providing a complete Java implementation with a demo that achieves near‑zero downtime during deployment.
This article explains how Kubernetes rolling deployments affect pod lifecycle, detailing the startup and shutdown phases, the impact of readiness probes and endpoint updates, and how to prevent connection interruptions by using preStop hooks and proper configuration for zero‑downtime deployments.
This article explains why and how to use Kubernetes features such as multiple pod replicas, PodDisruptionBudgets, deployment strategies, health probes, graceful termination, anti‑affinity, resource limits, and autoscaling to build zero‑downtime, highly available applications.
This guide explains how to achieve zero‑downtime releases of a Spring Boot application on Kubernetes by configuring readiness/liveness probes, rolling‑update strategies, graceful shutdown, horizontal pod autoscaling, Prometheus metrics collection, and externalized configuration via ConfigMaps.
This article analyzes why Kubernetes rolling updates can still cause service interruptions during pod startup and termination, explains the underlying mechanisms of Kubelet and Endpoint Controller, and provides practical steps such as readiness probes and preStop hooks to achieve smoother, near‑zero‑downtime deployments.
This article details the challenges, strategies, and step‑by‑step procedures for upgrading a 1,000‑node Kubernetes cluster from version 1.10 to 1.17 without service interruption, covering compatibility checks, in‑place versus replacement upgrades, container‑restart avoidance, pod eviction handling, and TCP connection issues.
Vivo’s internet team performed a zero‑downtime, in‑place upgrade of a 1,000‑node Kubernetes cluster from v1.10 to v1.17 by analyzing changelogs, backporting fixes, adjusting kubelet hash validation, adding tolerations, ensuring node labels, and using staged binary rollout, completing the process in roughly ten minutes.
The article explains why traditional three‑step application releases cause service interruptions, introduces three maturity levels for zero‑downtime deployment, compares blue‑green, rolling, and canary release models, and provides concrete technical components, load‑balancer architectures, and Spring‑Boot/Eureka shutdown procedures to achieve uninterrupted service.
This article describes how Coffee Meets Bagel upgraded a five‑node PostgreSQL 9.6 cluster to version 12.4 on AWS using pglogical logical replication, achieving less than 30 minutes of total downtime while sharing the architecture, migration steps, and lessons learned.
This article analyzes common database migration strategies, compares two approaches—master‑slave cut‑over and proxy‑based migration—describes Qunar's internal MMM and Percona XtraDB Cluster (PXC) architectures, and details a step‑by‑step, zero‑downtime migration process from MMM to PXC with practical lessons and best practices.
Lufax completed a two‑year, zero‑downtime, 100% Oracle‑to‑MySQL migration across its entire financial platform, detailing the dual‑write strategy, fine‑grained batch switching, tool‑driven automation, and the resulting cost and operational benefits for over 45 million users and supporting critical transaction services.
Xianyu’s zero‑downtime upgrade strategy for its multi‑microservice search stack combines rigorous forward/backward compatibility checks, batch‑wise stateless updates, careful stateful migrations via hot‑updates or dual‑write sync, robust service‑discovery traffic control, and Alibaba‑style monitoring, gray‑release testing, and rapid rollback to ensure uninterrupted service.
During the COVID‑19 surge, Tencent Cloud Redis scaled losslessly by using a proprietary slot‑specific RDB sync method that avoids downtime, handles large keys, Lua scripts, and multi‑key commands, outperforming open‑source and DTS approaches while minimizing resource duplication and maintaining 24/7 service.
This article examines classic smooth upgrade strategies for HTTP services, highlights their limitations, and provides a detailed analysis of Facebook’s graceful restart library for Go, including code examples, signal handling, process inheritance, and the mechanisms for zero‑downtime deployments.
This article explains how to achieve graceful, zero‑downtime restarts for long‑living TCP services by passing open file descriptors through a Unix domain socket, allowing new processes to inherit existing connections without interrupting clients, and provides a complete Go demo implementation.
Learn how to achieve zero‑downtime application upgrades by using Gunicorn’s HUP reload, uWSGI’s chained reload, load‑balancer node draining, and custom health‑check scripts that coordinate process restarts, ensuring minimal service interruption even for apps with long startup times.
The discussion explores practical strategies for adding an index to a MySQL table containing tens of millions of rows without stopping service, covering intermediate tables, double‑write syncing, rename operations, traffic considerations, lock types, and tooling such as Percona.