K8s_Meetup_Kitchener_N1_101

Transcript

1. Kubernetes Kitchener-Waterloo Welcome!
2. Thank you!
3. Page Introduction 3 Organizers of Toronto, Montreal, Ottawa, QC City, Kitchener-Waterloo meetups , a CloudOps initiative • Ayrat (Archy) Khayretdinov • Julia Simon
4. Page Kubernetes Kitchener-Waterloo - Plans for 2017-2018 • Community-driven - tell us what you want to learn! • Ecosystem • Quarterly meetups or more frequently if possible • Looking for sponsors, organizers, speakers (reach us via meetup.com) 4
5. http://K8scanadaslack.herokuapp.com Join the K8s Canada on Slack
6. Hands-on Vendor Agnostic Workshops Deepen your knowledge of containers and microservices and their ecosystems. • Docker and Kubernetes (2 days) • CI/CD (1 day) • Docker and Kubernetes Operations (2 days) • Kubernetes on Google Cloud (2 days) • Kubernetes on Azure (2 days) • Kubernetes on AWS (2 days) cloudops.com/docker-and-kubernetes-workshops info@cloudops.com
7. cloudops.com @cloudops_ Page Introductions: Instructor 7 Ayrat Khayretdinov • OpenStack & K8s Engineer and Technical Instructor at CloudOps • Kubernetes Openstack SIG • OpenStack-Helm • Organizer of K8s and Openstack meetups in Canada
8. Kitchener K8s Meetup Q4 • State of Container technologies • History of Kubernetes • Architecture Kubernetes • Kubernetes 101 • Workshop
9. Page I’m “Dockerized”, Now What? 9 Isolation: Keep jobs from interfering with each other Scheduling: Where should my job be run? Lifecycle: Keep my job running Discovery: Where is my job now? Constituency: Who is part of my job? Scale-up: Making my jobs bigger or smaller Auth{n,z}: Who can do things to my job? Monitoring: What’s happening with my job? Health: How is my job feeling?
10. cloudops.com @cloudops_ Is an ability to coordinate the containers in the clusters • When clusters consist of multiple nodes • When complex containerized applications are deployed Controls different aspects of containers lifecycle • Placement and initial deployment • Scaling and replication Allows moving from from a host-centric infrastructure to a container-centric infrastructure Container Orchestration Definition
11. cloudops.com @cloudops_ Container Orchestration Landscape
12. © 2015 Rancher Labs, Inc. 12
13. © 2015 Rancher Labs, Inc. 13
14. cloudops.com @cloudops_ Reason N1 Kubernetes Origin
15. Page 15 Kubernetes • Greek for “Helmsman”; also the root of the word “Governor” and “Cybernetics” • Container automation framework. • 100 %Open source, written in Go • K8s is an abbreviation derived by replacing the8letters "ubernete" with 8. Manage applications, not machines!
16. Page A Toolkit for Running Containerized Distributed Systems in Production 16 Co-locating helper processes Mounting storage systems Distributing secrets Application health checking Replicating application instances horizontal auto-scaling Naming and discovery Load balancing Rolling updates resource Monitoring Log access and ingestion support for debugging
17. Page Reason N1 Kubernetes Origin 17 Inspired by Borg … and later Omega from Google • Runs hundreds of thousands of jobs.. • ..for many thousands of different applications (YouTube, Search Index, Maps, etc.).. • ..across a number of clusters (hundreds).. • ..each with up to tens of thousands of machines.. • Borg currently manages many millions of physical servers! • ~50K machines / SRE • In production since 2003! • Offers a declarative job specification language, name service integration, real-time job monitoring, and tools to analyze and simulate system behavior.
18. cloudops.com @cloudops_ Reason N2 Production Adoption
19. cloudops.com @cloudops_ Pokemon GO (Gaming)
20. cloudops.com @cloudops_ Github (SAS based Software)
21. Ticketmaster - CoreOS Tectonic Summit 2016 Onsales = Black Friday every day! • Huge spikes / demand for tickets • Global company = across time zones • Limited inventory (Beyonce Tickets!) • Multiple sales channels 0 to 150M transactions in minutes! BIG SCALE, BIG CHALLENGES Solution: • K8S + Tectonic • Fully automated = 60 second app updates • High Confidence • Unlocked Daily Delivery Culture
22. cloudops.com @cloudops_
23. cloudops.com @cloudops_ Reason N3 Docker supports K8s
24. cloudops.com @cloudops_
25. Reason N4 Kubernetes Community & Ecosystem
26. Velocity 1.0 1.1 1.2 1.3 Total Commits Commits Since July 2014 1.4 1.5 1.6
27. Page 27 ~4k Commits in 1.5 +25% Unique Contributors Top 0.01% of all Github Projects 3500+ External Projects Based on K8s Companies Contributing Companies Using Community
28. Page 28
29. Page 29 Scalabilty v1.8 Goal: Support Large clusters & meet SLOs • Guaranteed SLO for <=5000 nodes • 150,000 podes supported • No more than 300 000 total containers • No more than 100 pods per node Note: Respecting <1s pod-startup policy
30. Page 30 Kubernetes Community • Special Interest Groups (SIGs) • Each SIG has a Lead, meetings and Slack channel • Release cycles ~ 3 months • Features go through Alpha, Beta, Stable cycle • Incubator: kubespray, bootkube and more • Graduated: Helm, Kompose
31. Kubecon 2017 Europe in Berlin
32. Kubecon 2017 North America Austin
33. Craig McLuckie Founder, CTO of Heptio Kubernetes key people Kelsey Hightower Developer advocate @Google Joe Beda Founder, CTO of Heptio Tim Hockin Developer @Google Brendan Burns Founder of kubernetes
34. Kelsey Hightower Kelsey Hightower Developer advocate
35. Kelsey Hightower Kelsey Hightower Developer advocate
36. Kelsey Hightower Kelsey Hightower Developer advocate
37. Key Players MONETIZATION STRATEGIES
38. Complementary Vendors and OpenSource tools NAMES WE HEAR, WHEN IT COMES TO… Deployment Configuration Management Storage Monitoring Networking Tectonic Kismatic Kubeadm Kops Kargo GKE ACS GiantSwarm StackPointCloud Terraform Ansible SaltStack Chef Cloudproviders GlusterFS Portworx Trident NetApp NFS Ceph/Rook Minio LongHorn Datagod Heapster Prometheus Sysdig Telegraf Hawkular Cloudproviders Calico Flannel Canal Weaveworks Envoy/Istio Contrail Contiv Romano
39. Kubernetes (K8s) 101 Architecture, Concepts and Features
40. Page Architecture 40
41. Page Container Runtime Interface (CRI) 1.7 - GA 41 AVOID LOCK-IN Goal of CRI: • Remove docker kubelet code of out Kubernetes • Simplify integration of K8s with other runtimes CRI runtimes • cri-docker • rktlet • cri-o (based on OCI) • cri-containerd (alpha) • virtlet (alpha) • frakti (alpha)
42. Page 42 Architecture
43. Page 43 Container Network Interface (CNI) 43 CNI is a specification proposed by CoreOS and adopted by Kubernetes. CNI is currnetly part of CNCF Goal of CNI: • To make network layer easy pluggable • CNM is not good option for K8s • Avoid code duplication Third-party CNI plugins: • flannel • weave • calico • Contiv and many more
44. Page 44 What about Windows ? 44
45. Page 45 Cloud Providers
46. Page Cloud Providers
47. Page Where you can run Kubernetes ? 47 • From Scratch • Public Cloud ◦ GCE, Amazon, Azure • Private Cloud ◦ OpenStack ◦ CloudStack • Bare-Metal • Hosted • On premises VMs ◦ VMware • Local-machine Solutions
48. How to use Kubernetes?
49. cloudops.com @cloudops_ Page Kubernetes CLI (kubectl) 49 Can be installed on any machine able to access the API server $ kubectl run echoserver \ --image=gcr.io/google_containers/echoserver:1.4 \ --port=8080 deployment "echoserver" created
50. cloudops.com @cloudops_ Page Kubernetes Web UI (Dashboard) 50 Dashboard UI allows to: • get an overview of applications running on the cluster • deploy containerized applications to a Kubernetes cluster • troubleshoot containerized applications • manage the cluster and its resources itself (Optional) Deployed as an Addon
51. cloudops.com @cloudops_ Page 51
52. Page Kubernetes Pod Manifest 52 apiVersion: v1 kind: Pod metadata: name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80 1 - kubernetes API version
53. Page Kubernetes Pod Manifest 53 apiVersion: v1 kind: Pod metadata: name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80 2 - Type of kubernetes object
54. Page Kubernetes Pod Manifest 54 apiVersion: v1 kind: Pod metadata: name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80 3 - Pod metadata (name, labels, annotations)
55. Page Kubernetes Pod Manifest 55 apiVersion: v1 kind: Pod metadata: name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80 4 - Pod specifications: (list of containers, volumes, ports etc.)
56. Page Example: Creating a Pod Create a pod: $ kubectl create -f nginx.yaml pod "nginx" created List all created pod: $ kubectl get pods NAME READY STATUS RESTARTS AGE nginx 1/1 Running 0 32s
57. Kubernetes Commercial Platforms Google Container Engine (GKE)
58. Kubernetes Cluter
59. cloudops.com @cloudops_ Page Kubernetes The Hard Way 59 Kelsey Hightower Developer advocate
60. cloudops.com @cloudops_ Page Kubernetes The Hard Way 60 1. Provisioning Compute Resources 2. Provisioning the CA and Generating TLS Certificates 3. Generating Kubernetes Configuration Files for Authentication 4. Generating the Data Encryption Config and Key 5. Bootstrapping the etcd Cluster 6. Bootstrapping the Kubernetes Control Plane 7. Bootstrapping the Kubernetes Worker Nodes 8. Configuring kubectl for Remote Access 9. Provisioning Pod Network Routes 10. Deploying the DNS Cluster Add-on
61. Page Minikube 61 Develop and test K8s apps locally Develop against local K8s Cluster Status: GA since Kubernetes 1.3 Check out: Minikube repo! Limitations: Cloud Provider features
62. cloudops.com @cloudops_ KubeAdm
63. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init
64. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init Kubernetes master initialized successfully! You can now join any number of nodes by running the following command: kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3
65. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init Kubernetes master initialized successfully! You can now join any number of nodes by running the following command: kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 node-01.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni node-01.myco.com# kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3
66. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init Kubernetes master initialized successfully! You can now join any number of nodes by running the following command: kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 node-01.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni node-01.myco.com# kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 Node join complete.
67. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init Kubernetes master initialized successfully! You can now join any number of nodes by running the following command: kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 node-01.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni node-01.myco.com# kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 Node join complete. master.myco.com# kubectl apply -f https://git.io/weave-kube
68. Steps: Simplified Setup master.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni master.myco.com# kubeadm init Kubernetes master initialized successfully! You can now join any number of nodes by running the following command: kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 node-01.myco.com# apt-get install -y kubelet kubeadm kubectl kubernetes-cni node-01.myco.com# kubeadm join --token 48b69e.b61e2d0dd5c 10.140.0.3 Node join complete. master.myco.com# kubectl apply -f https://git.io/weave-kube Network setup complete.
69. cloudops.com @cloudops_ Page KubeAdm Kubeadm is the official Kubernetes Project that allows to start Kubernetes cluster with 3 commands. The kubeadm tool is currently in alpha, however planned to be Beta soon. CONS • Infrastructure needs to be provisioned separately with Terrafom or similiar • Currently Alpha • Currently NON-HA • OpenStack cloudprovider experimental PROS • Advanced TLS BootStrap • All Kubernetes Features (RBAC, Calico, and etc.) • Always Up to date and close to upstream • Simple to use (3 commands to start a basic cluster ) • Runs anywhere • Can be further automated with (ansible, chef)
70. cloudops.com @cloudops_ KubeSpray incubated Kubernetes community project for deploying K8s clusters on premises or in the cloud.
71. cloudops.com @cloudops_ Page KubeSpray (ex. Kargo) 71 KubeSpray is a Kubernetes Incubator project that allows to deploy a Kubernetes cluster on Bare Metal, AWS, GCE and Openstack by using a combination of Terraform and Ansible playbooks.
72. cloudops.com @cloudops_ Page KubeSpray (ex. Kargo) 72 KubeSpray is a Kubernetes Incubator project that allows to deploy a Kubernetes cluster on Bare Metal, AWS, GCE and Openstack by using a combination of Terraform and Ansible playbooks. PROS • Capable of provisioning Openstack resources • Can deploy a Multi-Master/HA kubernetes cluster • Runs K8s cluster components as containers • Support most popular Linux distributions • Choose the network plugin to be used within the cluster: ◦ Flannel: gre/vxlan (layer 2) networking ◦ Weave: lightweight container overlay network ◦ Calico: bgp (layer 3) networking • Running latest K8s 1.4 • TLS secured endpoints. • Provides DNS Addon integration CONS • KubeSpray is a great tool for advanced users. Can be used for integration with automation tools since KubeSpray provides a CLI only for deployment of K8s cluster. Therefore it is not very suitable for users or developers who are not comfortable with Ansible, Openstack env. and KubeSpray config file which has huge number of options.
73. Page KOPS (Kubernetes Operations) - kubectl for K8s clusters. Kops lets you deploy production-grade, highly available, Kubernetes clusters from the command line. Deployment is currently supported on Amazon Web Services (AWS), with more platforms planned. KOPs 73
74. Kubernetes Concepts
75. Pods
76. Page Concept: Pods 76 Pod Git Synchronizer Node.js App Container Volume Consumers git Repo Small group of containers & volumes Logical application Containers within a pod are tightly coupled Shared namespaces • Containers in a pod share IP, port and IPC namespaces • Containers in a pod talk to each other through localhost Ephemeral
77. Page Kubernetes Pod Manifest 77 apiVersion: v1 kind: Pod metadata: name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80
78. Page Example: Creating a Pod Create a pod: $ kubectl create -f nginx.yaml pod "nginx" created List all created pod: $ kubectl get pods NAME READY STATUS RESTARTS AGE nginx 1/1 Running 0 32s
79. DB API Frontend Labels
80. Page Kubernetes Pod Manifest with Label 80 apiVersion: v1 kind: Pod metadata: labels: app: frontend name: nginx spec: containers: - name: nginx image: nginx:1.7.9 ports: - containerPort: 80
81. DB API Frontend Deployments
82. Page Deployment - (Replication Controller v2) 82 82 ... Defines a desired state for logical group of Pods • Replica Sets (RS) Features • Replication • Rolling upgrade
83. cloudops.com @cloudops_ Replication
84. Page Deployments (Replication Controllers V2) 84 node 1 f0118 node 3 node 4 node 2 d9376 b0111 a1209 Replication Controller - Desired = 4 - Current = 4
85. Page Deployments (Replication Controllers V2) 85 node 1 f0118 node 3 node 4 node 2 d9376 b0111 a1209 Replication Controller - Desired = 4 - Current = 4
86. Page Deployments (Replication Controllers V2) 86 node 1 f0118 node 3 node 4 b0111 a1209 Replication Controller - Desired = 4 - Current = 3
87. Page Deployments (Replication Controllers V2) 87 node 1 f0118 node 3 node 4 c9bad a1209 Replication Controller - Desired = 4 - Current = 4 b0111
88. cloudops.com @cloudops_ Rolling Update
89. Page Feature: Rolling Update Node1 Node3 Node2 ghost Pod app v1 Service ghost Pod app v1 Pod app v1
90. Page Feature: Rolling Update Node1 Node3 Node2 ghost Pod app v1 Service ghost Pod app v1 Pod app v1 Pod app v2
91. Page Feature: Rolling Update Node1 Node3 Node2 ghost Pod app v1 Service ghost Pod app v1 Pod app v1 Pod app v2
92. Page Feature: Rolling Update Node1 Node3 Node2 ghost Pod app v1 Service ghost Pod app v1 Pod app v1 Pod app v2
93. Page Feature: Rolling Update Node1 Node3 Node2 Service ghost Pod app v1 Pod app v1 Pod app v2
94. Page Feature: Rolling Update Node1 Node3 Node2 Service ghost Pod app v1 Pod app v1 Pod app v2 Pod app v2
95. Page Feature: Rolling Update Node1 Node3 Node2 Service ghost Pod app v1 Pod app v1 Pod app v2 Pod app v2
96. Page Feature: Rolling Update Node1 Node3 Node2 Service ghost Pod app v1 Pod app v1 Pod app v2 Pod app v2
97. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v1 Pod app v2 Pod app v2
98. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v1 Pod app v2 Pod app v2 Pod app v2
99. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v1 Pod app v2 Pod app v2 Pod app v2
100. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v1 Pod app v2 Pod app v2 Pod app v2
101. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v1 Pod app v2 Pod app v2 Pod app v2
102. Page Feature: Rolling Update Node1 Node3 Node2 Service Pod app v2 Pod app v2 Pod app v2
103. Page Example: Defining a Replication Controller template: metadata: name: nginx labels: app: nginx spec: containers: - name: nginx image: nginx ports: - containerPort: 80 apiVersion: apps/v1beta2 kind: Deployment metadata: name: nginx spec: replicas: 3 selector: app: nginx template: ... Pod template
104. DB API Frontend Services
105. Page Concept: Services 105 • A group of pods that act as one == Service – group == selector • Services Types: – Load balanced (from CloudProvider) – NodePort (Minikube) – ClusterIP (interal only) – Ingress (HTTP/ HTTPS) Portal (VIP) Client Service - Name = “nifty-svc” - Selector = {“App”: “Nifty”} - Port = 9376 - ContainerPort = 8080 Portal IP is assigned
106. cloudops.com @cloudops_ Example: Defining a Service apiVersion: v1 kind: Service metadata: name: my-service spec: ports: - port: 80 protocol: TCP targetPort: 9376 selector: app: MyApp
107. cloudops.com @cloudops_ Example: Service Operations Create a service $ kubectl create -f myservice.yaml List all services on a cluster $ kubectl get services Delete a service $ kubectl delete service myservice
108. DB API Frontend Services
109. cloudops.com @cloudops_ • Liveness • Readiness • ConfigMaps • Secrets • Volumes • Jobs and CronJobs • Init Containers Kubernetes of Kubernetes • Stateful Sets • Autoscaling • Network Policies • RBAC • Federation • Helm
110. 1.2 • Deployments • DaemonSets • New UI • Ingress • TPR • Improved Scale 1.3 • Legacy application support • Cluster Federation • More nodes • In-cluster IAM • Scheduled jobs • Cluster autoscaling • Public cloud dashboard
111. 1.4 • Kubeadm • Extended stateful apps • Cluster Fed improvements • Dashboard improvements • Pod Security Policy 1.5 • Introduction of CRI (Alpha) • Stateful Sets (Beta) • OpenAPI support • Support for Windows • Server-based containers (Alpha)
112. 1.6 • 5000 Nodes clusters • RBAC (Alpha) • Controlled scheduling • Dynamic Storage Provisioning 1.7 • Encrypting Secrets in etcd • Stateful Set Upgrades • CRD • Local Storage (alpha)
113. • RBAC [Stable] • Network Policy Egress Traffic [Beta] • Spark on Kubernetes [Beta] • Volume Resize and Snapshots 1.8