An Automated and Riskless Solution to Deliver Resources Accross Multiple...- Yikun Wang & Jun Zhang, Ant Group
蚂蚁集团跨多个集群交付资源的自动化和无风险解决方案 | An Automated and Riskless Solution to Deliver Resources Accross Multiple Clusters in Ant Group - Yikun Wang & Jun Zhang, Ant Group
部署的滚动更新是一个巧妙的设计,在Kubernetes生态系统中的一些项目也提供了交付能力。然而,在生产和金融场景中这些还不够。在蚂蚁集团,这些需求包括:使用git作为唯一的真相源;多维度和自适应策略;跨多个集群的编排;支持大规模集群;以及自动自愈的能力。基于以上原因,蚂蚁集团提供了一套Rollout套件。开发人员在git中维护多维度策略。在提交提交时,Rollout和SelfHeal Controller将观察工作负载并触发渐进式交付,同时对失败的任务进行自愈。引入Controller Mesh的能力来支持大规模集群的水平扩展操作员。本演示将描述相关原因和架构,并为您提供所有核心能力的详细信息。
The Deployment's rolling update is a clever design, and some projects in Kubernetes ecosystem also provide delivery capabilities. However, these are not enough in production and financial scenarios. At Ant Group, these requirements include: using git as the single source of truth; multi-dimensional and adaptive strategies; orchestration acrossing multiple clusters; support for large-scale clusters; and the ability to automatically self-heal. For the reasons above, Ant Group has provided a set of Rollout suites. developers maintain multi-dimensional strategies in git. When submitting a commit, the Rollout and SelfHeal Controller will observe workloads and trigger a gradient delivery while also performing self-healing for failed tasks. The ability of Controller Mesh is introduced to scale operators horizontally for supporting large-scale clusters. This presentation will describe the relevant causes and architecture, and provide you with detailed information on all core capabilities.
CNCF概况(幻灯片)
扫描二维码联系我们!
CNCF (Cloud Native Computing Foundation)成立于2015年12月,隶属于Linux Foundation,是非营利性组织。
CNCF(云原生计算基金会)致力于培育和维护一个厂商中立的开源生态系统,来推广云原生技术。我们通过将最前沿的模式民主化,让这些创新为大众所用。请关注CNCF微信公众号。