Shulou(Shulou.com)06/02 Report--

OAM, the world's first open application model, is open source.

On October 17, 2019, Jiang Jiangwei (nickname: Xiaoxie), partner of Alibaba and general manager of Aliyun Intelligent basic products Division, announced in Qcon Shanghai that Aliyun and Microsoft jointly launched the open application model Open Application Model (OAM) open source project. OAM's vision is to communicate and connect application developers, operators, and application infrastructure in a standardized way, making the management and delivery of cloud native applications more concise, efficient, and controllable.

KubeVirt enters CNCF Sandbox

KubeVirt despite the convenience provided by container technology, it is still inevitable to use VM in certain situations (editor's note: VM is just one of the options, and different secure container solutions can replace simple VM in different ways). The KubeVirt project from Redhat provides the ability to prepare, deploy, run, and manage VM within a K8s cluster, allowing users to use a VM in the same way as Pod. Now, this project has officially entered the Sandbox of CNCF.

Megalix releases KubeAdvisor version 1.0

KubeAdvisor released version 1.0 a few days ago. KubeAdvisor, a tool to assist K8s operation and maintenance, can scan the resources, status and configuration of K8s cluster, and provide cluster users with the most valuable information to help ensure the reliability and stability of infrastructure and upper applications by providing appropriate "observability" (editor's note: dazzling monitoring data and graphs are not equal to observability).

Important upstream progress

1. Allows you to dynamically adjust the log level of kube-scheduler. Kube-scheduler is an important default scheduling component of K8s. In some cases, their logs can reflect a lot of information. This PR allows the log printing level of kube-scheduler to be adjusted dynamically.

two。 A recent move on the scheduler side is to Score Plugin some Prioritizing logic. Several recent related PR are listed below, which can be followed by interested students.

BalancedResourceAllocationMostRequestedPriorityLeastResourcePriority

3. Add a metric to scheduler that tracks the number of goroutine for Binding and Prioritizing.

In some special scenarios such as large-scale application deployment, the scheduler may become the bottleneck of the overall performance.

4. Here are two PR that optimize the kubeadm experience:

Kubeadm structured output: structured output of some kubeadm output information (yaml, json, etc.), easy to continue processing; kubeadm added-- with-ca flag to display ca expiration information: ca expiration is one of the common willingness to cause components to be unavailable, with this flag can more easily guide ca expiration information.

5. Here are some important / interesting KEP that interested friends can follow:

Added graduation conditions for the beta version of Kube API Server's network proxy. KAS allows network traffic to configure Kube API Server to (or less than) the specified proxy;insecure kubelet log: through a switch, if the serving cert of kubelet expires (kube apiserver does not know kubelet) but kube-apiserver 's client cert does not expire (kubelet knows kube apiserver), kubelet is allowed to return some log by skipping tls verification. This feature is useful in testing and debug scenarios. Standardize the implementation of conformanece test content (documentation, API schema, code review, expert knowledge, etc.) extend NodeRestriction Controller to limit more operations that Node can Pod (mainly from security considerations); the status of HPA reaches implementable

The two planned GA are good entry points for novice contributors who want to participate in the K8s community (requirements with clear goals and methods).

By tagging Node with taint according to Node condition, automation helps the scheduler identify unsuitable scheduling nodes; previously, the scheduling logic of DaemonSet is in DaemonSet controller, and this KEP wants to move the scheduling logic to the scheduler. VMware-Tanzu is recommended for open source projects

VMware has begun to fully support K8s, and one of its recent open source initiatives is to migrate several of its own cloud native open source projects to the new Organization:VMware-Tanzu. This project currently includes the following projects:

Velero: application migration tool octant: a cluster status display dashboardsonobuoy: a K8s analysis tool k8s-transmogrifier

A large amount of API is abandoned in K8s 1.16, which affects a lot of cluster configuration and Helm chart that have been used in production. Here is a tool to automatically convert the API of depreciated in K8s 1.16, which can be found out by those who need it.

Read the recommendations this week

1. "how did Alibaba's R & D model evolve? "

With the continuous development of cloud computing, many developers are interested in the changes that this technology will bring to the way of development. Cloud computing solves the problem of the transition from CAPEX to OPEX. What are the tangible benefits of the cloud? How should application architecture be done in a cloud environment? Based on his personal experience from traditional architecture to cloud architecture, Jiang Jiangwei (Xiaoxie), partner of Alibaba and general manager of Aliyun Intelligent basic products, explained the value points brought by the new architecture and new R & D model. This article is based on the speech delivered by Jiang Jiangwei (Xiaoxie) in QCon Shanghai 2019.

2. "How Zalando manages 140 + Kubernetes Cluster"

This article introduces some practical experience gained by Zalando's team in managing a large number of K8s clusters (on the public cloud), such as always deploying dual clusters (prod & non-prod) for each domain or production community, using Github managed configuration files, managing upgrades through CLM (Cluster Lifecycle Manager), and so on.

3. "Liveness Probes are Dangerous"

Liveness Probe and Readiness Probe are important tools in K8s to determine whether an application is available or not, but there are risks associated with incorrect use of Liveness Probe or Readiness Probe (such as accidental use of external dependencies, Liveness Probe sometimes prevents Pod from correctly entering a failed state and cannot fully recover from health, etc.). This article summarizes the best practices for using both.

4. "A Practical Guide to Setting Kubernetes Requests and Limits"

Request and limit in the definition of Kubernetes resources is a clich é. In addition to better explaining these concepts, this article also provides suggestions on how to configure them from the perspective of SLA.

5. Protecting Kubernetes API Against CVE-2019-11253 (Billion Laughs Attack) and Other Vulnerabilities

By introducing an example in K8s that may lead to billions laughs attacks vulnerabilities, this article explains the general security operations of using K8s in a production environment, including correctly configuring RBAC, periodically checking Role and RoleBinding, never exposing the address of Master host, and so on.

6. "implementation of Weather Service based on Knative Serverless Technology-part two"

In the last issue, we introduced how to implement weather service based on Knative Serverless technology-in the last article, we introduced how to interface with Knative to store event sources, subscribe and send weather reminders through the channel service provided by table storage. For more wonderful articles about Knative, please see the Knative series of articles.

The original link to this article is the original content of Yunqi community and may not be reproduced without permission.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.