“Human error is not our cause of troubles; instead, human error is a consequence of the design of the tools that we gave them.” 2. Accidents are due to the inevitable design problems in complex systems that TO AUTOMATE AND CAPTURE ORGANIZATIONAL KNOWLEDGE i. ChatOps pioneered at GitHub – put automation tools (Hubot) into the middle of their chatrooms 1. Everyone saw everything that was happening 2. New infrastructure, and environments 2. Deployment tools 3. Testing standards and tools, including security 4. Deployment pipeline tools 5. Monitoring and analysis tools 6. Tutorials and standards ii. Google

checked into version control 1. Application code & dependencies 2. Environment scripts & creation tools 3. DB scripts and reference data 4. Containers 5. Automated tests 6. Project artifacts – documentation 7. Application configuration files 8. This also includes pre-production and build processes 9. Tools iv. 2014 State of DevOps Report – use of version control by Ops was the highest predictor of both Demonstrated in a production-like environment iv. Ideally, use the same monitoring, logging, and other tools in pre-production environments as in production v. Dev & Ops gain shared mastery of application

so that new development does not cause expensive break-fix activity.  It has good monitoring tools in place.  It is coded in a way that resists hard-to-find defects like concurrency errors. Build  The developer uses a powerful integrated development environment (IDE) that centralizes the tools that help him or her do the job.  The code is built automatically using Continuous Integration automatically through scripted tests.  It is deployed automatically with automated deployment tools onto infrastructure that can be automatically provisioned. The choice we have to make:  Option

visitors move through my siteLinux Performance Tools by Brendan Gregg (Netflix) @brendangregg http://www.slideshare.net/brendangregg/velocity-2015-linux-perf-tools  Value of mentoring (vs self-teaching through

create telemetry for functions built ii. Generate graphs with overlays of production changes iii. Tools – StatsD, JMX, codahale, New Relic, Prometheus, etc. f. CREATE SELF-SERVICE ACCESS TO TELEMETRY AND contributing to the cause of the issue 2. Focus on dynamically registered & discovered supporting tools rather than static 3. Graph business metrics with infrastructure metrics to provide deeper context

OF WORK i. Keep the current state visible and up to date ii. Use tools to reinforce desired behavior - anthropologists describe tools as a cultural artifact iii. Create shared backlogs and reporting

crio） ★安装crictl工具 默认随kubeadm一起安装了，在cri-tools-xxxx.x86_64.rpm这个包里，如果想单独 安装，则直接装这个rpm包或者到github上下载此软件 # wget h�ps://github.com/kubernetes-sigs/cri- tools/releases/download/v1.28.0/crictl-v1.28.0-linux-amd64 name: busybox #容器名，小写字母开头，后可接数字-减号， 不可带小数点 image: cof-lee.com/tools/busybox:latest #容器使用的镜像 imagePullPolicy: IfNotPresent #镜像获取方式，优先本地，后网络拉

Do start a book club. Book List  Scaling Lean & Agile Development: Thinking & Organizational Tools for Large-Scale Scrum, Craig Larman  Continuous Delivery: Reliable Software Releases through Build

actions.  Automate as much of the development and deployment activities as possible.  Adopt common tools from planning and requirements through deployment and operations.  Leverage agile software principles