This is a creation in Article, where the information may have evolved or changed.
What is the difference between Apache's Mesos and Google's kubernetes? This article comes from a question on the StackOverflow, mainly discusses the difference between Mesos and kubernetes, I believe many of us also have the question of agreeing. Kubernetes's developer Craig answered the question, while Masi also made an overview, not necessarily for the reader's reference. Kubernetes is primarily for container clusters, and Mesos is suitable for any framework and application, so kubernetes can run on Mesos.
Kubernetes is an open source project that introduces Google's cluster management tools to virtual machines and bare-metal scenarios. It works perfectly in modern operating system environments (such as CoreOS and Red Hat Atomic) and provides lightweight compute nodes that you can control. The kubernetes is developed using Golang and is lightweight, modular, portable, and extensible. We (the Kubernetes development team) are working with a number of different technology companies, including mesosphere that maintain the Mesos project, to upgrade kubernetes to a standard way of interacting with compute clusters. Kubernetes has re-implemented Google's experience in building cluster applications. These concepts include the following:
- Pods: A method of organizing containers together;
- Replication Controllers: A method of controlling the life cycle of a container (translator Note: Replication controller ensures that at any time a specified number of pod copies (replicas) are running in the Kubernetes cluster);
- Labels: A way to find and query containers;
- Services: A group of containers used to implement a particular function;
As a result, you can start, migrate, and scale your cluster quickly and easily with kubernetes. In this case, the cluster is like a virtual machine-like flexible resource, it is a logical unit of operations. Open it, use it, resize it, and then close it, just so fast, it's that simple.
Mesos and kubernetes have similar visions, but they have different advantages over different life cycles. Mesos is a distributed system kernel that integrates different machines on a single logical computer. Mesos comes in handy when you have a lot of physical resources and want to build a huge, static computing cluster. There are many modern scalable data processing applications that can be run on Mesos, including Hadoop, Kafka, Spark, and so on, while you can use container technology to run all your data processing applications in a pool of underlying resources. In some ways, Mesos is a more heavyweight project than Kubernetes, but thanks to Mesosphere-like contributors, Mesos is becoming simpler and easier to manage.
Interestingly, Mesos is accepting the idea of kubernetes and has begun to support Kubernetes API. So if you need them, it will be a convenient way to get more power for your kubernetes application (such as the high-availability backbone, more advanced dispatch commands, the ability to control a large number of nodes), It is also well suited to the product-level work environment (after all, Kubernetes is still an initial release).
When asked the difference, I will answer:
- If you are a novice in a clustered world, that Kubernetes is a great start. It can solve your problem in the quickest, simplest, and most lightweight way possible, and help you with cluster-oriented development. It provides a high level of portability, as many vendors have started to support kubernetes such as Microsoft, IBM, Red Hat, CoreOS, mesosphere, VMware, etc.
- If you have tasks that already exist (Hadoop, Spark, Kafka, and so on), then Mesos can provide you with a framework to interleave different tasks, and then add something new, such as a kubernetes application.
- If you want to use the function Kuberntes has not been implemented, that Mesos is a good substitute, after all, it is ripe.
Original link: Whats The difference between Apaches Mesos and Googles kubernetes (translation: Liu Kaining proofreading: Li Yingjie)
from:http://dockerone.com/article/147