Downloaded 22 times
Uploaded byBen Hall
Experimenting and Learning Kubernetes and Tensorflow
This document discusses experimenting with Kubernetes and Tensorflow. It begins with an introduction of the author and overview of learning via interactive browser-based labs on Katacoda.com. Then it demonstrates setting up Minikube and Kubeadm to create Kubernetes clusters, deploying Tensorflow models and services on Kubernetes using Deployments and Jobs, and considerations for scaling Kubernetes and Tensorflow workloads. It concludes with a call to action for others to share their experiences by writing scenarios on Katacoda.
More Related Content
![Cloud Computing Bootcamp On The Google App Engine [v1.1]](https://cdn.slidesharecdn.com/ss_thumbnails/cloudcomputingbootcamponthegoogleappenginev1-1-090506155059-phpapp02-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
Similar to Experimenting and Learning Kubernetes and Tensorflow
![[Spark Summit 2017 NA] Apache Spark on Kubernetes](https://cdn.slidesharecdn.com/ss_thumbnails/apachesparkonkubernetespublic1-170929072840-thumbnail.jpg?width=600ounds&width=560&fit=bounds)
Experimenting and Learning Kubernetes and Tensorflow
- 1. Experimenting and Learning Kubernetesand Tensorflow @Ben_Hall Ben@BenHall.me.uk Katacoda.com
- 4. Experimenting and Learning Kubernetesand Tensorflow @Ben_Hall Ben@BenHall.me.uk Katacoda.com
- 5.
- 6. Learn via InteractiveBrowser-Based Labs Katacoda.com
- 7. • In thenext 25/30 minutes • Learning to Learn • Creating Kubernetes Experiment Playground • Running Tensorflow on Kubernetes • Keeping up to date with the community
- 9.
- 13.
- 14.
- 15.
- 17.
- 18.
- 19.
- 20.
- 22.
- 23. apiVersion: extensions/v1beta1 kind: Deployment metadata: name:inception-deployment labels: k8s-app: inception-deployment spec: replicas: 3 selector: matchLabels: k8s-app: inception-deployment template: metadata: labels: k8s-app: inception-deployment annotations: scheduler.alpha.kubernetes.io/critical-pod: '' spec: containers: - name: inception-container image: katacoda/tensorflow_serving imagePullPolicy: Never command: - /bin/sh - -c args: - /serving/bazel-bin/tensorflow_serving/model_servers/tensorflow_model_server --port=9000 --model_name=inception --model_base_path=/serving/inception-export ports: - containerPort: 9000
- 25. apiVersion: batch/v1 kind: Job metadata: name:inception-client spec: template: metadata: name: inception-client spec: containers: - name: inception-client image: katacoda/tensorflow_serving imagePullPolicy: Never command: - /bin/bash - -c args: - /serving/bazel-bin/tensorflow_serving/example/inception_client --server=inception-deployment:9000 --image=/data/cat.jpg volumeMounts: - name: inception-persistent-storage mountPath: /data volumes: - name: inception-persistent-storage hostPath: path: /root restartPolicy: Never
- 26. Kubernetes and Tensorflowat scale?
- 27.
- 28.
- 29. Worker Worker WorkerWorker Parameter Server Parameter Server
- 30. # On ps0.example.com: $python trainer.py --ps_hosts=ps0.example.com:2222,ps1.example.com:2222 --worker_hosts=worker0.example.com:2222,worker1.example.com:2222 --job_name=ps --task_index=0 # On ps1.example.com: $ python trainer.py --ps_hosts=ps0.example.com:2222,ps1.example.com:2222 --worker_hosts=worker0.example.com:2222,worker1.example.com:2222 --job_name=ps --task_index=1 # On worker0.example.com: $ python trainer.py --ps_hosts=ps0.example.com:2222,ps1.example.com:2222 --worker_hosts=worker0.example.com:2222,worker1.example.com:2222 --job_name=worker --task_index=0 # On worker1.example.com: $ python trainer.py --ps_hosts=ps0.example.com:2222,ps1.example.com:2222 --worker_hosts=worker0.example.com:2222,worker1.example.com:2222 --job_name=worker --task_index=1
- 31. Kubernetes namespace Kubernetes PodKubernetesPodKubernetes Pod Containerized TF Worker Containerized TF Worker Containerized TF Worker Kubernetes Deployment Containerized TF PS Server 1 Server 2 Server 3 Kubernetes namespace Kubernetes PodKubernetes PodKubernetes Pod Containerized TF Worker Containerized TF Worker Containerized TF Worker Kubernetes Deployment Containerized TF PS Storage Server 4 Server 5
- 32. Kubernetes namespace Kubernetes PodKubernetesPodKubernetes Pod Containerized TF Worker Containerized TF Worker Containerized TF Worker Kubernetes Deployment Containerized TF PS GPU1 GPU2 GPU3 Server 1 Server 2 GPU1 GPU2 GPU3 Server 3 Kubernetes namespace Kubernetes PodKubernetes PodKubernetes Pod Containerized TF Worker Containerized TF Worker Containerized TF Worker Kubernetes Deployment Containerized TF PS Storage Server 4 Server 5
- 33. Docker Container andGPU docker run -it --device /dev/nvidia0:/dev/nvidia0 --device /dev/nvidia1:/dev/nvidia1 --device /dev/nvidiactl:/dev/nvidiactl --device /dev/nvidia-uvm:/dev/nvidia-uvm tf-cuda:v1.1beta /bin/bash
- 38. Summary • Kubernetes isdesigned for running distributed systems at scale • The model of Tensorflow fits cleanly into Kubernetes • As Tensorflow usage increases, Kubernetes can scale to meet demands
- 39.
- 40. Call To Action •Interested in sharing your Kubernetes or Tensorflow experience? Write your own scenarios and teach interactively! • Teaching teams internally? Private Katacoda environments
- 41.