Chaos engineering platform chaosblade box new heavy release

author ： Mingshao

What is chaos Engineering

The system architecture has experienced from stand-alone to distributed , And now the cloud native architecture , Its complexity is increasing , The difficulty of problem location also rises . In the face of faults that may occur at any time , Is there any way to solve this dilemma .

Chaos Engineering （Chaos Engineering） The discipline of conducting experiments on distributed systems , By actively injecting faults , Identify weaknesses in the system ahead of time , Promote the improvement of Architecture , Ultimately, business resilience . So as to avoid failure in the online operation environment .

Here is an example of cloud native architecture , Why can chaos engineering solve the problems existing in the system architecture . The correspondence between cloud native architecture principle and chaos engineering principle can be found , Explain with the principle of service , The fundamental principle of service is how to manage services , That is, the problem of judging the strong and weak dependence between upstream and downstream Services . Through chaos Engineering , You can locate the request to a specific machine , Then reduce to the application of specific machines , Constantly minimize the explosion radius , By injecting faults between applications , Judge whether the upstream and downstream services are normal , To judge its strong and weak dependence .

The goal of chaos engineering is to achieve a resilient architecture , There are two parts here ： Ductile system and ductile structure . Resilient systems have redundancy 、 Extensibility 、 Immutable infrastructure 、 Stateless application 、 Avoid cascading failures . Resilient organizations include efficient delivery 、 Failure plan 、 Emergency response mechanism . Highly resilient systems can also have unexpected failures , So the tough organization can make up for the missing part of the tough system , Building the ultimate resilience architecture through chaos Engineering .

Chaos engineering is a way of injecting faults actively , Identify weaknesses in the system ahead of time , Promote architecture improvement , Ultimately, business resilience . For people with different functions, chaos engineering is introduced , Its business value is different ：

• Architects ： It can help them verify the fault tolerance of the architecture

• Development / Operation and maintenance ： It can improve the emergency efficiency of its failure

• test ： Help them expose online problems in advance , Reduce the failure recurrence rate

• product / Design ： Remind the customer of the use experience

How to implement chaos Engineering

For enterprises or businesses, how to implement chaos Engineering ？ Is there any tool or platform that can help it land quickly ？

ChaosBlade  It is a tool of chaos experiment execution that follows chaos experiment model , It has high scene richness , Simple and easy to use , Support for multiple platforms 、 Multilingual environment , Include  Linux、Kubernetes  and  Docker  platform , Support  Java、NodeJS、C++、Golang  Language application . Support  200  Multiple scenes ,3000  Multiple parameters . It is a fault injection tool for the end side , But when the business is launched , There will be the following problems ：

• How to visualize the fault injection process ？

• How to inject faults into multiple clusters or hosts at the same time ？

• How to get the statistics of the overall drill

• ......

So in  ChaosBlade  It also needs a platform layer , Manage and rehearse the implementation tools of chaotic Engineering .

ChaosBlade-Box  It is oriented to multiple clusters 、 Multilingual 、 Multiple environments , Open source cloud native chaos engineering console .

The overall architecture of the open source platform and injection tool is as follows , It mainly includes several constituent modules ：

• ChaosBlade-Box Console ： Chaos experiment user interface

• ChaosBlade-Box：Server  The back-end service , It mainly includes the arrangement of drill scenes and security control 、 Chaos engineering tool deployment （ChaosBlade、LitmusChaos...）、 Support probe management and multi-dimensional experiments

• Agent： probe , There are mainly （ChaosBlade-Box）Server Connect and keep your heart beating 、 Report  K8s  related data 、 Exercise functions such as command distribution channel

• ChaosBlade： Deployed on the business host or  K8s  Within cluster , Tools for drilling on the end side

new edition  ChaosBlade-Box  The platform is a multi cluster oriented 、 Multiple environments 、 Multi language cloud native chaos engineering platform . Support international Chinese English switching , Support global namespace , So that the same user can according to their own needs , Set different global namespaces , Such as ： Test space 、 Sandbox space and online space . Provide automated tool deployment , Simplify tool installation steps , Improve execution efficiency . The platform supports probe installation and drilling in different environments , Such as host and  Kubernetes, among  Kubernetes  Support in the environment  Node、Pod、Container  Drill under dimension . stay  Kubernetes  In the environment, it will automatically collect  Pod  related data , And unified management in application management , This simplifies the user's query steps , There is no need to go to the cluster to view the applications to be rehearsed  Pod  Name or  Container  name . And support one click migration to enterprise , Synchronize the drill data of the community version to the enterprise version on demand .

The following is in the new version  ChaosBlade-Box  The whole process of a drill on the platform , Support sequential execution 、 Stage performs two process choreography , Sequential execution means that multiple drill scenarios take effect in turn , The value of stage execution is that multiple drill scenarios take effect at the same time . Ensure that the drill is restored through a variety of security strategies , Such as manual punishment and automatic stop , Automatic stop is configured by setting the timeout parameter during the drill configuration , So even if the platform and probe （Agent） Out of contact , When manual stop is not possible , Also when the timeout time arrives , Automatic recovery of faults .

What are the advantages of the new version

Compared with the old version , The front-end interface is unified with the enterprise version , Simplify the switching cost of using habits , More perfect international Chinese English switching , And support global namespace switching ; The back end provides a smoother rehearsal , Perfect application management , And strengthened the control of the probe , And support one click migration to enterprise ; The function of the probe is strengthened , It provides a more perfect  API, It supports multi environment deployment and can be used as a drill channel in different environments , Support automatic installation and uninstallation , And collect and report data to simplify the drill .

Related links

Middleware developer conference address （ Speech  PDF  Downloadable ）：

https://developer.aliyun.com/topic/middleware/developer/summit

MSE  First purchase of professional edition of registration configuration center  9  A discount ,MSE  Full specification of cloud native gateway prepaid  85  A discount .