How to Install and Use Kubernetes by Weaveworks
Download as PPTX, PDF1 like873 views
This document provides an overview of how to install and use Kubernetes. It discusses key Kubernetes concepts like pods, deployments, services and how they relate. It also summarizes the Kubernetes architecture and components. The presentation encourages attendees to join the Weave user group for more training on continuous delivery, monitoring and network policy in Kubernetes.
How to Install and Use Kubernetes by Weaveworks
- 1. How to install and use Kubernetes Jeff Hoffer, Developer Experience github.com/eudaimos
- 2. What does Weave do? Weave helps devops iterate faster with: • observability & monitoring • continuous delivery • container networks & firewalls Kubernetes is our #1 platform
- 3. Agenda 1. Concepts 2. Containers, Pods, Deployments, Services 3. Installing Kubernetes 4. Demos of Pods, Deployments, Services 5. Microservices sample app 6. What’s next?
- 4. Kubernetes: all you need to know Pods containers Services Deployments
- 10. web Concepts Podnginx apiVersion: v1 kind: Pod metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 Node
- 11. Recap: all you need to know Pods containers Container Image Docker container image, contains your application code in an isolated environment. Pod A set of containers, sharing network namespace and local volumes, co- scheduled on one machine. Mortal. Has pod IP. Has labels.
- 12. Motivation for Deployments Node 1 Pod web nginx Node 2
- 13. Motivation for Deployments Pod web nginx Node 1 Node 2
- 14. Motivation for Deployments Pod web nginx Node 1 Node 2
- 15. Deployment web nginx apiVersion: extensions/v1beta1 kind: Deployment metadata: name: nginx-deployment spec: replicas: 1 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 Node 1
- 16. Deployment web nginx apiVersion: extensions/v1beta1 kind: Deployment metadata: name: nginx-deployment spec: replicas: 1 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 Node 1
- 20. Deployment web nginx apiVersion: extensions/v1beta1 kind: Deployment metadata: name: nginx-deployment spec: replicas: 1 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 Node 1
- 21. Deployment web nginx apiVersion: extensions/v1beta1 kind: Deployment metadata: name: nginx-deployment spec: replicas: 2 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 Node 1
- 23. Recap: all you need to know Pods containers Deployments Container Image Docker container image, contains your application code in an isolated environment. Pod A set of containers, sharing network namespace and local volumes, co- scheduled on one machine. Mortal. Has pod IP. Has labels. Deployment Specify how many replicas of a pod should run in a cluster. Then ensures that many are running across the cluster.
- 24. Service discovery • Kubernetes provides DNS for service discovery
- 25. Service discovery • Kubernetes provides DNS for service discovery
- 26. Services: ClusterIP (internal things) Computer 1 web ruby Computer 2 10.0.0.1 db pgsql 10.0.0.2 service VIP 10.1.0.1DNS lookup “db” returns A 10.1.0.1
- 27. Services: ClusterIP (internal things) Computer 1 web ruby Computer 2 10.0.0.1 db pgsql 10.0.0.2 service VIP 10.1.0.1
- 28. Services: NodePort (external) Computer 1 web nginx web nginx 10.0.0.1 10.0.0.2 81.23.64.18 81.23.64.19 requests requests Computer 2
- 29. Computer 1 web nginx web nginx 10.0.0.1 10.0.0.2 81.23.64.18 81.23.64.19 requests requests Computer 2 Services: NodePort (external) NodePort 30001 service VIP 10.1.0.1 NodePort 30001 service VIP 10.1.0.1
- 30. Computer 1 web nginx web nginx 10.0.0.1 10.0.0.2 81.23.64.18 81.23.64.19 requests requests Computer 2 Services: NodePort (external) NodePort 30001 service VIP 10.1.0.1 NodePort 30001 service VIP 10.1.0.1
- 31. kind: Deployment metadata: name: nginx-deployment spec: replicas: 2 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 kind: Service metadata: name: frontend spec: type: NodePort selector: app: nginx ports: - port: 80 targetPort: 80 nodePort: 30002 Using selectors How do services connect to deployments?
- 32. kind: Deployment metadata: name: nginx-deployment spec: replicas: 2 template: metadata: labels: app: nginx spec: containers: - name: nginx image: nginx:1.7.9 kind: Service metadata: name: frontend spec: type: NodePort selector: app: nginx ports: - port: 80 targetPort: 80 nodePort: 30002 How do you expose services to outside? Using NodePort
- 33. Recap: all you need to know Pods containers ServicesDeployments Container Image Docker container image, contains your application code in an isolated environment. Pod A set of containers, sharing network namespace and local volumes, co- scheduled on one machine. Mortal. Has pod IP. Has labels. Deployment Specify how many replicas of a pod should run in a cluster. Then ensures that many are running across the cluster. Has labels. Service Names things in DNS. Gets virtual IP. Two types: ClusterIP for internal services, NodePort for publishing to outside. Routes based on labels.
- 34. Architecture of Kubernetes itself Node 1 Node 2 Master
- 35. API server Architecture of Kubernetes itself Node 1 Node 2 Master API server etcd kubeadm init
- 36. API server Architecture of Kubernetes itself Node 1 Node 2 Master API server etcd kubeadm init kubeadm join kubelet
- 37. API server Architecture of Kubernetes itself Node 1 Node 2 Master API server etcd kubeadm init kubeadm join kubeadm join kubelet kubelet
- 38. API server Architecture of Kubernetes itself Node 1 Node 2 Master API server etcd kubeadm init kubeadm join kubeadm join kubelet kubelet kubectl apply
- 39. API server Architecture of Kubernetes itself Node 1 Node 2 Master containers Services containers API server etcd kubeadm init kubeadm join kubeadm join kubectl apply kubelet kubelet
- 40. Training!
- 41. Join the Weave user group! meetup.com/pro/Weave/ weave.works/help
- 42. What’s next? Continuous delivery: hooking up my CI/CD pipeline to Kubernetes How do I monitor this stuff? Network policy for security Come to our Weave Cloud training to find out!
- 43. Thanks! Questions? We are hiring! DX in San Francisco Engineers in London & SF weave.works/weave-company/hiring
Editor's Notes
- #10: Containers in a pod share a network namespace; pods have IP addresses So “localhost” is the same. They can also share access to volumes
- #11: any questions?
- #15: as tim hockin likes to say, pods are mortal if a machine goes away where a pod is running, nothing tries to restart it
- #17: looks like a pod
- #21: deployments can also help you with scaling
- #23: any questions?
- #27: services are how things can talk to each other and also how the outside world can talk to things kube-proxy creates iptables NAT rules so that packets going to a VIP get routed to a set of service endpoints where set of endpoints is a subset of pod IP addresses matching the label selector, it only includes pods that are in a ready state
- #28: services are how things can talk to each other and also how the outside world can talk to things kube-proxy creates iptables NAT rules so that packets going to a VIP get routed to a set of service endpoints where set of endpoints is a subset of pod IP addresses matching the label selector, it only includes pods that are in a ready state
- #29: services are how things can talk to each other and also how the outside world can talk to things kube-proxy creates iptables NAT rules so that packets going to a VIP get routed to a set of service endpoints where set of endpoints is a subset of pod IP addresses matching the label selector, it only includes pods that are in a ready state
- #30: services are how things can talk to each other and also how the outside world can talk to things kube-proxy creates iptables NAT rules so that packets going to a VIP get routed to a set of service endpoints where set of endpoints is a subset of pod IP addresses matching the label selector, it only includes pods that are in a ready state
- #31: services are how things can talk to each other and also how the outside world can talk to things kube-proxy creates iptables NAT rules so that packets going to a VIP get routed to a set of service endpoints where set of endpoints is a subset of pod IP addresses matching the label selector, it only includes pods that are in a ready state
- #33: you can also use LoadBalancer if you have set up cloud provider integration, and it will auto provision load balancers from e.g. AWS
- #41: https://www.katacoda.com/courses/weave/kubernetes-training
- #42: https://www.katacoda.com/courses/weave/kubernetes-training
- #43: https://www.katacoda.com/courses/weave/kubernetes-training