{"id":19211995,"url":"https://github.com/joedborg/juju-k8s-tf","last_synced_at":"2026-03-02T13:04:15.052Z","repository":{"id":145095595,"uuid":"105919911","full_name":"joedborg/Juju-K8s-TF","owner":"joedborg","description":"Guide for Juju -\u003e Kubernetes -\u003e Tensorflow + GPU","archived":false,"fork":false,"pushed_at":"2017-10-05T17:59:33.000Z","size":7,"stargazers_count":0,"open_issues_count":0,"forks_count":0,"subscribers_count":1,"default_branch":"master","last_synced_at":"2025-11-14T13:24:56.106Z","etag":null,"topics":[],"latest_commit_sha":null,"homepage":null,"language":"Smarty","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/joedborg.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2017-10-05T17:33:15.000Z","updated_at":"2017-10-05T17:59:34.000Z","dependencies_parsed_at":null,"dependency_job_id":"908d7465-5a73-4ab8-b290-63a2d25d0d9f","html_url":"https://github.com/joedborg/Juju-K8s-TF","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/joedborg/Juju-K8s-TF","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joedborg%2FJuju-K8s-TF","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joedborg%2FJuju-K8s-TF/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joedborg%2FJuju-K8s-TF/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joedborg%2FJuju-K8s-TF/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/joedborg","download_url":"https://codeload.github.com/joedborg/Juju-K8s-TF/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/joedborg%2FJuju-K8s-TF/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":30003489,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-03-02T12:19:43.414Z","status":"ssl_error","status_checked_at":"2026-03-02T12:19:02.215Z","response_time":60,"last_error":"SSL_connect returned=1 errno=0 peeraddr=140.82.121.6:443 state=error: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":[],"created_at":"2024-11-09T13:45:23.251Z","updated_at":"2026-03-02T13:04:15.001Z","avatar_url":"https://github.com/joedborg.png","language":"Smarty","funding_links":[],"categories":[],"sub_categories":[],"readme":"# Tensorflow on Kubernetes on AWS with GPU support\n## Prerequisites\n* Installed `juju`.\n* Installed `kubectl`.\n* Installed `helm`.\n* Access to `p2.xlarge` and `m4.xlarge` instances in `eu-central-1`.\n  * This requires contacting AWS support.\n  * At the time of writing, `eu-central-1` was the only region I could get these\ninstances.\n\n## Method\n### Juju\n1) Bootstrap `juju bootstrap aws/eu-central-1`.\n2) Deploy `juju deploy ./k8s-tensorflow.yaml`.\n### EFS\n3) Setup an EFS drive in `eu-central-1` and assign the correct security groups.  (See https://medium.com/intuitionmachine/kubernetes-gpus-tensorflow-8696232862ca#c678).\n4)\n  a) `cd ./charts`.\n  b) `cp efs/values.yaml ./efs.yaml`.\n  c) Update `efs.yaml` to the correct id.\n  d) Deploy with `helm install efs --name efs --values ./efs.yaml`.\n  e) `cd ../`.\n### CNN\n5)\n  a) `cd ./charts`.\n  b) `cp distributed-cnn/values.yaml ./dcnn.yaml`.\n  c) Get ingress IP of the k8s cluster and append XIP DNS with `echo $(juju status kubernetes-worker-cpu --format json | jq -r '.machines[].\"dns-name\"').xip.io`.\n  d) Update `dcnn.yaml` to correct the DNS field with the output of the\n  previous command.\n  e) Deploy with `helm install distributed-cnn --name dcnn --values ./dcnn.yaml`.\n  f) Browse to XIP DNS address to see TensorBoard.\n\n## Notes\n* The CNN example does not facilitate the GPUs.  Roll you own for this (See https://medium.com/intuitionmachine/kubernetes-gpus-tensorflow-8696232862ca#eb05).\n* Renting these instances is mega expensive (around $100 a day), be prepared for\n  this!\n* If you have issues with the pods at step `5`, the likely issue is that the\n  EFS volumes doesn't have the correct security groups assigned.\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjoedborg%2Fjuju-k8s-tf","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fjoedborg%2Fjuju-k8s-tf","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fjoedborg%2Fjuju-k8s-tf/lists"}