Skip to content

Instantly share code, notes, and snippets.

@fjudith

fjudith/h2o-kubeflow_guide_fre.md

Last active November 12, 2018 16:40
Download ZIP

Revisions

  1. fjudith revised this gist Nov 12, 2018. 1 changed file with 0 additions and 12 deletions.
    12 changes: 0 additions & 12 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -232,16 +232,6 @@ ks apply ${KF_ENV} -c katib
    # ks apply ${KF_ENV} -c spartakus
    ```

    ```bash
    ks apply ${KF_ENV} -c ambassador
    ks apply ${KF_ENV} -c jupyterhub
    ks apply ${KF_ENV} -c centraldashboard
    ks apply ${KF_ENV} -c tf-job-operator
    ks apply ${KF_ENV} -c argo
    ks apply ${KF_ENV} -c katib
    ks apply ${KF_ENV} -c spartakus
    ```

    Valider que tous les Pods sont opérationnels

    ```bash
    @@ -346,9 +336,7 @@ ks prototype use io.ksonnet.pkg.h2o3-static h2o3-static \
    --cpu 1 \
    --replicas 2 \
    --model_server_image ${H2O3_IMAGE}
    ```

    ```bash
    ks apply ${KF_ENV} -c h2o3-static -n ${K8S_NAMESPACE}
    ```

  2. fjudith revised this gist Nov 12, 2018. 1 changed file with 26 additions and 8 deletions.
    34 changes: 26 additions & 8 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -384,38 +384,56 @@ Exécuter le code source suivant, en cliquant sur le bouton `Run` à chaque sect
    ```python
    import h2o
    from h2o.automl import H2OAutoML
    h2o.init(ip='h2o3-static', port='54321')
    h2o.init(ip='h2o3-static.kubeflow.svc.cluster.local', port='54321')

    train_loc = 'https://storage.googleapis.com/h2o-bq-large-dataset/higgs_train_10k.csv'
    test_loc = 'https://storage.googleapis.com/h2o-bq-large-dataset/higgs_test_5k.csv'

    train = h2o.import_file(train_loc)
    test = h2o.import_file(test_loc)
    # Import a sample binary outcome train/test set into H2O
    train = h2o.import_file("https://s3.amazonaws.com/erin-data/higgs/higgs_train_10k.csv")
    test = h2o.import_file("https://s3.amazonaws.com/erin-data/higgs/higgs_test_5k.csv")
    ```

    ```python
    # Identify predictors and response
    x = train.columns
    y = "response"
    x.remove(y)
    ```

    ```python
    # For binary classification, response should be a factor
    train[y] = train[y].asfactor()
    test[y] = test[y].asfactor()
    ```

    aml = H2OAutoML(max_runtime_secs = 60)
    ```python
    # Run AutoML for 30 seconds
    aml = H2OAutoML(max_runtime_secs = 30)
    aml.train(x = x, y = y,
    training_frame = train)
    ```

    ```python
    # View the AutoML Leaderboard
    lb = aml.leaderboard
    lb
    ```

    ```python
    preds = aml.predict(test)
    # The leader model is stored here
    aml.leader
    ```

    ```python
    # If you need to generate predictions on a test set, you can make
    # predictions directly on the `"H2OAutoML"` object, or on the leader
    # model object directly

    preds = aml.predict(test)
    print(preds.shape)
    preds
    ```
    ```python
    # or:
    preds = aml.leader.predict(test)
    print(preds.shape)
    preds
    ```
  3. fjudith revised this gist Nov 12, 2018. 1 changed file with 44 additions and 22 deletions.
    66 changes: 44 additions & 22 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -245,31 +245,53 @@ ks apply ${KF_ENV} -c spartakus
    Valider que tous les Pods sont opérationnels

    ```bash
    kubectl get po,svc,pvc --namespace ${NAMESPACE}
    kubectl get po,svc,pvc --namespace ${K8S_NAMESPACE}
    ```

    ```text
    NAME READY STATUS RESTARTS AGE
    ambassador-5798c6798f-k4n76 2/2 Running 1 21h
    ambassador-5798c6798f-lxz2w 2/2 Running 1 21h
    ambassador-5798c6798f-wzw5c 2/2 Running 1 21h
    tf-hub-0 1/1 Running 0 20h
    tf-job-operator-845d5686d4-nhw6k 1/1 Running 0 21h
    ```

    Valider que les services sont bien en écoute.

    ```bash
    kubectl get svc --namespace "${K8S_NAMESPACE}"
    ```

    ```text
    NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
    ambassador ClusterIP 10.3.0.244 <none> 80/TCP 21h
    ambassador-admin ClusterIP 10.3.0.75 <none> 8877/TCP 21h
    k8s-dashboard ClusterIP 10.3.0.52 <none> 443/TCP 21h
    tf-hub-0 ClusterIP None <none> 8000/TCP 21h
    tf-hub-lb ClusterIP 10.3.0.154 <none> 80/TCP 21h
    NAME READY STATUS RESTARTS AGE
    pod/ambassador-c97f7b448-5qjpk 3/3 Running 0 28m
    pod/ambassador-c97f7b448-x47nk 3/3 Running 0 28m
    pod/ambassador-c97f7b448-z45kt 3/3 Running 0 28m
    pod/argo-ui-7495b79b59-vsjbg 1/1 Running 0 27m
    pod/centraldashboard-798f8d68d5-w77hv 1/1 Running 0 28m
    pod/modeldb-backend-d69695b66-sg6nf 1/1 Running 0 21m
    pod/modeldb-db-975db58f7-gn4sh 1/1 Running 0 21m
    pod/modeldb-frontend-78ccff78b7-2wgkm 1/1 Running 0 22m
    pod/studyjob-controller-7df5754ddf-l6pmh 1/1 Running 0 21m
    pod/tf-hub-0 1/1 Running 0 28m
    pod/tf-job-dashboard-7499d5cbcf-hjp2w 1/1 Running 0 27m
    pod/tf-job-operator-v1alpha2-644c5f7db7-k8lps 1/1 Running 0 27m
    pod/vizier-core-56dfc85cf9-sg2fk 1/1 Running 1 21m
    pod/vizier-db-6bd6c6fdd5-dczfk 1/1 Running 0 21m
    pod/vizier-suggestion-bayesianoptimization-5d5bc5685c-pvbpx 1/1 Running 0 21m
    pod/vizier-suggestion-grid-5dbfc65587-2rmlk 1/1 Running 0 21m
    pod/vizier-suggestion-hyperband-5d9997fb99-56w9t 1/1 Running 0 21m
    pod/vizier-suggestion-random-7fccb79977-z6s2t 1/1 Running 0 21m
    pod/workflow-controller-d5cb6468d-6lvhj 1/1 Running 0 27m
    NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
    service/ambassador ClusterIP 10.3.0.234 <none> 80/TCP 28m
    service/ambassador-admin ClusterIP 10.3.0.71 <none> 8877/TCP 28m
    service/argo-ui NodePort 10.3.0.220 <none> 80:41378/TCP 50m
    service/centraldashboard ClusterIP 10.3.0.7 <none> 80/TCP 28m
    service/k8s-dashboard ClusterIP 10.3.0.180 <none> 443/TCP 28m
    service/modeldb-backend ClusterIP 10.3.0.227 <none> 6543/TCP 27m
    service/modeldb-db ClusterIP 10.3.0.23 <none> 27017/TCP 27m
    service/modeldb-frontend ClusterIP 10.3.0.77 <none> 3000/TCP 27m
    service/statsd-sink ClusterIP 10.3.0.243 <none> 9102/TCP 28m
    service/tf-hub-0 ClusterIP None <none> 8000/TCP 50m
    service/tf-hub-lb ClusterIP 10.3.0.171 <none> 80/TCP 50m
    service/tf-job-dashboard ClusterIP 10.3.0.22 <none> 80/TCP 50m
    service/vizier-core NodePort 10.3.0.32 <none> 6789:30678/TCP 22m
    service/vizier-db ClusterIP 10.3.0.235 <none> 3306/TCP 27m
    service/vizier-suggestion-bayesianoptimization ClusterIP 10.3.0.196 <none> 6789/TCP 27m
    service/vizier-suggestion-grid ClusterIP 10.3.0.201 <none> 6789/TCP 27m
    service/vizier-suggestion-hyperband ClusterIP 10.3.0.91 <none> 6789/TCP 27m
    service/vizier-suggestion-random ClusterIP 10.3.0.81 <none> 6789/TCP 27m
    NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
    persistentvolumeclaim/vizier-db Bound pvc-98ac1da6-e66a-11e8-b904-96000012e4d2 10Gi RWO rook-ceph-block 27m
    ```

    ### Déployer le volume persistent pour le notebook H2O
  4. fjudith revised this gist Nov 12, 2018. 1 changed file with 3 additions and 3 deletions.
    6 changes: 3 additions & 3 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -260,7 +260,7 @@ tf-job-operator-845d5686d4-nhw6k 1/1 Running 0 21h
    Valider que les services sont bien en écoute.

    ```bash
    kubectl get svc --namespace "${NAMESPACE}"
    kubectl get svc --namespace "${K8S_NAMESPACE}"
    ```

    ```text
    @@ -319,15 +319,15 @@ Exécuter les commandes suivantes pour initialiser le serveur H2O3.
    ```bash
    ks prototype use io.ksonnet.pkg.h2o3-static h2o3-static \
    --name h2o3-static \
    --namespace kubeflow \
    --namespace ${K8S_NAMESPACE} \
    --memory 1 \
    --cpu 1 \
    --replicas 2 \
    --model_server_image ${H2O3_IMAGE}
    ```

    ```bash
    ks apply ${KF_ENV} -c h2o3-static -n kubeflow
    ks apply ${KF_ENV} -c h2o3-static -n ${K8S_NAMESPACE}
    ```

    ## Déployer le Notebook H2O
  5. fjudith revised this gist Nov 12, 2018. 1 changed file with 49 additions and 16 deletions.
    65 changes: 49 additions & 16 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -166,37 +166,70 @@ export GITHUB_TOKEN=
    ```

    ```bash
    KUBERNETES_VERSION=v1.11.2
    KUBEFLOW_VERSION=v0.3.2
    APPNAME="kubeflow-demo"
    KUBEFLOW_REGISTRY="github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"
    H2O_REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    NAMESPACE=kubeflow
    KF_ENV=local
    H2O3_IMAGE="quay.io/fjudith/h2o3"

    kubectl create namespace ${NAMESPACE}

    KUBERNETES_VERSION=${KUBERNETES_VERSION:-"v1.11.2"}
    KUBEFLOW_VERSION=${KUBEFLOW_VERSION:-"v0.3.2"}
    APPNAME=${APPNAME:-"kubeflow-demo"}
    KUBEFLOW_REGISTRY=${KUBEFLOW_REGISTRY:-"github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"}
    H2O_REGISTRY=${H2O_REGISTRY:-"github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"}
    K8S_NAMESPACE=${K8S_NAMESPACE:-kubeflow}
    KF_ENV=${KF_ENV:-local}
    H2O3_IMAGE=${H2O3_IMAGE:-"quay.io/fjudith/h2o3"}

    kubectl create namespace ${K8S_NAMESPACE}

    # create workspace
    mkdir -p ~/h2oworkspace
    pushd ~/h2oworkspace

    # create the ksonnet app
    pushd ~/h2oworkspace
    ks init ${APPNAME} --api-spec=version:${KUBERNETES_VERSION}
    cd ~/h2oworkspace/${APPNAME}

    # remove the default environment; The cluster might not exist yet
    # so we might be pointing to the wrong cluster.
    ks env rm default

    # add ksonnet registry to app containing all the kubeflow manifests as maintained by Google Kubeflow team
    ks registry add kubeflow ${KUBEFLOW_REGISTRY}
    # add ksonnet registry to app containing all the h2o component manifests
    ks registry add h2o-kubeflow ${H2O_REGISTRY}

    # install components from kubeflow and h2o3-kubeflow registries
    ks pkg install kubeflow/argo@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/core@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/examples@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/katib@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/mpi-job@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/pytorch-job@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/seldon@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/tf-serving@${KUBEFLOW_VERSION}
    #ks pkg install kubeflow/tf-job@${KUBEFLOW_VERSION}
    ks pkg install h2o-kubeflow/h2o3-static

    ks generate core kubeflow-core --name="kubeflow-core" --namespace="${NAMESPACE}"
    ks env add ${KF_ENV} --api-spec=version:${KUBERNETES_VERSION}
    # Required only for GKE ks param set kubeflow-core cloud gke --env=cloud

    # generate all required components
    ks generate pytorch-operator pytorch-operator
    ks generate ambassador ambassador
    ks generate jupyterhub jupyterhub
    ks generate centraldashboard centraldashboard
    ks generate tf-job-operator tf-job-operator
    ks generate argo argo
    ks generate katib katib

    # remove anonymous usage reporting enabled using spartakus
    #local usageId=$(((RANDOM<<15)|RANDOM))
    #ks generate spartakus spartakus --usageId=${usageId} --reportUsage=true

    ks env add ${KF_ENV} --api-spec=version:${KUBERNETES_VERSION} --namespace "${K8S_NAMESPACE}"
    # required only for GKE ks param set kubeflow-core cloud gke --env=cloud

    # deploy kubeflow
    ks apply ${KF_ENV} -c ambassador
    ks apply ${KF_ENV} -c jupyterhub
    ks apply ${KF_ENV} -c centraldashboard
    ks apply ${KF_ENV} -c tf-job-operator
    ks apply ${KF_ENV} -c argo
    ks apply ${KF_ENV} -c katib
    # ks apply ${KF_ENV} -c spartakus
    ```

    ```bash
  6. fjudith revised this gist Nov 12, 2018. 1 changed file with 8 additions and 2 deletions.
    10 changes: 8 additions & 2 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -200,13 +200,19 @@ ks env add ${KF_ENV} --api-spec=version:${KUBERNETES_VERSION}
    ```

    ```bash
    ks apply ${KF_ENV} --component="kubeflow-core"
    ks apply ${KF_ENV} -c ambassador
    ks apply ${KF_ENV} -c jupyterhub
    ks apply ${KF_ENV} -c centraldashboard
    ks apply ${KF_ENV} -c tf-job-operator
    ks apply ${KF_ENV} -c argo
    ks apply ${KF_ENV} -c katib
    ks apply ${KF_ENV} -c spartakus
    ```

    Valider que tous les Pods sont opérationnels

    ```bash
    kubectl get pods --namespace ${NAMESPACE}
    kubectl get po,svc,pvc --namespace ${NAMESPACE}
    ```

    ```text
  7. fjudith revised this gist Nov 12, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -166,7 +166,7 @@ export GITHUB_TOKEN=
    ```

    ```bash
    KUBERNETES_VERSION=v1.12.2
    KUBERNETES_VERSION=v1.11.2
    KUBEFLOW_VERSION=v0.3.2
    APPNAME="kubeflow-demo"
    KUBEFLOW_REGISTRY="github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"
  8. fjudith revised this gist Nov 12, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -191,7 +191,7 @@ ks registry add h2o-kubeflow ${H2O_REGISTRY}
    # install components from kubeflow and h2o3-kubeflow registries
    ks pkg install kubeflow/core@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/tf-serving@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/tf-job@${KUBEFLOW_VERSION}
    #ks pkg install kubeflow/tf-job@${KUBEFLOW_VERSION}
    ks pkg install h2o-kubeflow/h2o3-static

    ks generate core kubeflow-core --name="kubeflow-core" --namespace="${NAMESPACE}"
  9. fjudith revised this gist Nov 12, 2018. 1 changed file with 4 additions and 4 deletions.
    8 changes: 4 additions & 4 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -104,7 +104,7 @@ ubuntu config --default-user root
    Lancer Ubuntu, puis exécuter les commandes suivantes pour installer l'outil de déploiement d'applications Kubernetes `ks` (ksonnet).

    ```bash
    KSONNET_VERSION=0.12.0
    KSONNET_VERSION=0.13.0

    wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VERSION}/ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
    tar -zvxf ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
    @@ -118,7 +118,7 @@ wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VER
    Depuis la session de commande **Ubuntu**, exécuter les commandes suivantes pour installer le client d'administration Kubernetes `kubectl`.

    ```
    KUBECTL_VERSION=1.11.2
    KUBECTL_VERSION=1.12.2
    curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl && \
    chmod +x ./kubectl && \
    @@ -166,8 +166,8 @@ export GITHUB_TOKEN=
    ```

    ```bash
    KUBERNETES_VERSION=v1.11.2
    KUBEFLOW_VERSION=v0.2.5
    KUBERNETES_VERSION=v1.12.2
    KUBEFLOW_VERSION=v0.3.2
    APPNAME="kubeflow-demo"
    KUBEFLOW_REGISTRY="github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"
    H2O_REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
  10. fjudith revised this gist Oct 23, 2018. 1 changed file with 0 additions and 8 deletions.
    8 changes: 0 additions & 8 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -332,14 +332,6 @@ train = h2o.import_file(train_loc)
    test = h2o.import_file(test_loc)
    ```

    ```python
    x = train.columns
    y = "response"
    x.remove(y)
    train[y] = train[y].asfactor()
    ```


    ```python
    x = train.columns
    y = "response"
  11. fjudith revised this gist Sep 10, 2018. 1 changed file with 4 additions and 4 deletions.
    8 changes: 4 additions & 4 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -104,7 +104,7 @@ ubuntu config --default-user root
    Lancer Ubuntu, puis exécuter les commandes suivantes pour installer l'outil de déploiement d'applications Kubernetes `ks` (ksonnet).

    ```bash
    KSONNET_VERSION=0.10.1
    KSONNET_VERSION=0.12.0

    wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VERSION}/ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
    tar -zvxf ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
    @@ -118,7 +118,7 @@ wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VER
    Depuis la session de commande **Ubuntu**, exécuter les commandes suivantes pour installer le client d'administration Kubernetes `kubectl`.

    ```
    KUBECTL_VERSION=1.10.2
    KUBECTL_VERSION=1.11.2
    curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl && \
    chmod +x ./kubectl && \
    @@ -166,8 +166,8 @@ export GITHUB_TOKEN=
    ```

    ```bash
    KUBERNETES_VERSION=v1.10.1
    KUBEFLOW_VERSION=v0.1.2
    KUBERNETES_VERSION=v1.11.2
    KUBEFLOW_VERSION=v0.2.5
    APPNAME="kubeflow-demo"
    KUBEFLOW_REGISTRY="github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"
    H2O_REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
  12. fjudith revised this gist May 21, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -128,7 +128,7 @@ sudo mv ./kubectl /usr/local/bin/kubectl
    ### Initialiser kubectl

    ```bash
    pushd ~/git/coreos-kubernetes/multi-node/vagrant && PATH=${PATH}:$(pwd) && source init-kubectl.shpu && popd
    pushd ~/git/coreos-kubernetes/multi-node/vagrant && PATH=${PATH}:$(pwd) && source init-kubectl.sh && popd
    ```

    ## Télécharger le notebook Jupyter H2O
  13. fjudith revised this gist Apr 29, 2018. 1 changed file with 20 additions and 8 deletions.
    28 changes: 20 additions & 8 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -163,9 +163,14 @@ Depuis la session de commande **Ubuntu**. Lancer les commandes suivantes pour in
    # Github API token with repo/public_repo permissions
    export GITHUB_TOKEN=
    #
    ```

    ```bash
    KUBERNETES_VERSION=v1.10.1
    KUBEFLOW_VERSION=v0.1.2
    APPNAME="kubeflow-demo"
    REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    KUBEFLOW_REGISTRY="github.com/kubeflow/kubeflow/tree/${KUBEFLOW_VERSION}/kubeflow"
    H2O_REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    NAMESPACE=kubeflow
    KF_ENV=local
    H2O3_IMAGE="quay.io/fjudith/h2o3"
    @@ -175,21 +180,26 @@ kubectl create namespace ${NAMESPACE}
    mkdir -p ~/h2oworkspace
    pushd ~/h2oworkspace

    ks init ${APPNAME} --api-spec=version:v1.10.0
    ks init ${APPNAME} --api-spec=version:${KUBERNETES_VERSION}
    cd ~/h2oworkspace/${APPNAME}

    # add ksonnet registry to app containing all the kubeflow manifests as maintained by Google Kubeflow team
    ks registry add kubeflow https://github.com/kubeflow/kubeflow/tree/master/kubeflow
    ks registry add kubeflow ${KUBEFLOW_REGISTRY}
    # add ksonnet registry to app containing all the h2o component manifests
    ks registry add h2o-kubeflow ${REGISTRY}
    ks pkg install kubeflow/core
    ks pkg install kubeflow/tf-serving
    ks pkg install kubeflow/tf-job
    ks registry add h2o-kubeflow ${H2O_REGISTRY}

    # install components from kubeflow and h2o3-kubeflow registries
    ks pkg install kubeflow/core@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/tf-serving@${KUBEFLOW_VERSION}
    ks pkg install kubeflow/tf-job@${KUBEFLOW_VERSION}
    ks pkg install h2o-kubeflow/h2o3-static

    ks generate core kubeflow-core --name="kubeflow-core" --namespace="${NAMESPACE}"
    ks env add ${KF_ENV}
    ks env add ${KF_ENV} --api-spec=version:${KUBERNETES_VERSION}
    # Required only for GKE ks param set kubeflow-core cloud gke --env=cloud
    ```

    ```bash
    ks apply ${KF_ENV} --component="kubeflow-core"
    ```

    @@ -275,7 +285,9 @@ ks prototype use io.ksonnet.pkg.h2o3-static h2o3-static \
    --cpu 1 \
    --replicas 2 \
    --model_server_image ${H2O3_IMAGE}
    ```

    ```bash
    ks apply ${KF_ENV} -c h2o3-static -n kubeflow
    ```

  14. fjudith revised this gist Apr 29, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -175,7 +175,7 @@ kubectl create namespace ${NAMESPACE}
    mkdir -p ~/h2oworkspace
    pushd ~/h2oworkspace

    ks init ${APPNAME}
    ks init ${APPNAME} --api-spec=version:v1.10.0
    cd ~/h2oworkspace/${APPNAME}

    # add ksonnet registry to app containing all the kubeflow manifests as maintained by Google Kubeflow team
  15. fjudith revised this gist Apr 28, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -104,7 +104,7 @@ ubuntu config --default-user root
    Lancer Ubuntu, puis exécuter les commandes suivantes pour installer l'outil de déploiement d'applications Kubernetes `ks` (ksonnet).

    ```bash
    KSONNET_VERSION=0.9.2
    KSONNET_VERSION=0.10.1

    wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VERSION}/ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
    tar -zvxf ks_${KSONNET_VERSION}_linux_amd64.tar.gz && \
  16. fjudith revised this gist Apr 28, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -118,7 +118,7 @@ wget --quiet https://github.com/ksonnet/ksonnet/releases/download/v${KSONNET_VER
    Depuis la session de commande **Ubuntu**, exécuter les commandes suivantes pour installer le client d'administration Kubernetes `kubectl`.

    ```
    KUBECTL_VERSION=1.10.0
    KUBECTL_VERSION=1.10.2
    curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl && \
    chmod +x ./kubectl && \
  17. fjudith revised this gist Apr 24, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -43,7 +43,7 @@ choco install conemu docker docker-compose docker-machine docker-machine-vmwarew
    * [ ] Télécharger le notebook jupyter H2O
    * [ ] Installer H2O-Kubeflow
    * [ ] Déployer le Notebook H2O
    * [x] Validation
    * [ ] Validation

    ## Télécharger la Box Vagrant CoreOS

  18. fjudith revised this gist Apr 24, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -43,7 +43,7 @@ choco install conemu docker docker-compose docker-machine docker-machine-vmwarew
    * [ ] Télécharger le notebook jupyter H2O
    * [ ] Installer H2O-Kubeflow
    * [ ] Déployer le Notebook H2O
    * [ ] Validation
    * [x] Validation

    ## Télécharger la Box Vagrant CoreOS

  19. fjudith revised this gist Apr 20, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -167,7 +167,7 @@ export GITHUB_TOKEN=
    APPNAME="kubeflow-demo"
    REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    NAMESPACE=kubeflow
    KF_ENV=default
    KF_ENV=local
    H2O3_IMAGE="quay.io/fjudith/h2o3"

    kubectl create namespace ${NAMESPACE}
  20. fjudith revised this gist Apr 20, 2018. 1 changed file with 2 additions and 1 deletion.
    3 changes: 2 additions & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -170,6 +170,8 @@ NAMESPACE=kubeflow
    KF_ENV=default
    H2O3_IMAGE="quay.io/fjudith/h2o3"

    kubectl create namespace ${NAMESPACE}

    mkdir -p ~/h2oworkspace
    pushd ~/h2oworkspace

    @@ -185,7 +187,6 @@ ks pkg install kubeflow/tf-serving
    ks pkg install kubeflow/tf-job
    ks pkg install h2o-kubeflow/h2o3-static

    kubectl create namespace ${NAMESPACE}
    ks generate core kubeflow-core --name="kubeflow-core" --namespace="${NAMESPACE}"
    ks env add ${KF_ENV}
    # Required only for GKE ks param set kubeflow-core cloud gke --env=cloud
  21. fjudith revised this gist Apr 20, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -167,7 +167,7 @@ export GITHUB_TOKEN=
    APPNAME="kubeflow-demo"
    REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    NAMESPACE=kubeflow
    KF_ENV=local
    KF_ENV=default
    H2O3_IMAGE="quay.io/fjudith/h2o3"

    mkdir -p ~/h2oworkspace
  22. fjudith revised this gist Apr 18, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -310,7 +310,7 @@ Exécuter le code source suivant, en cliquant sur le bouton `Run` à chaque sect
    ```python
    import h2o
    from h2o.automl import H2OAutoML
    h2o.init(ip='h2o3', port='54321')
    h2o.init(ip='h2o3-static', port='54321')

    train_loc = 'https://storage.googleapis.com/h2o-bq-large-dataset/higgs_train_10k.csv'
    test_loc = 'https://storage.googleapis.com/h2o-bq-large-dataset/higgs_test_5k.csv'
  23. fjudith revised this gist Apr 18, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -13,7 +13,7 @@ https://blog.h2o.ai/2018/03/h2o-kubeflow-kubernetes-how-to/

    Minimum | Recommandé
    ------- | ----------
    **4 virtualcore** = 2 CPU-core 16 GB RAM + Hyperthreading | **8 Virtual core**: 4-core 32GB RAM + Hyperthreading
    **4 Virtual core** = 2 CPU-core 16 GB RAM + Hyperthreading | **8 Virtual core**: 4-core 32GB RAM + Hyperthreading
    Window 10 Fall Creator Update | Window 10 Fall Creator Update
    Kubernetes cluster 1.9+ [coreos-kubernetes](https://github.com/fjudith/coreos-kubernetes) |
    `1 controller (1GB), 1 etcd (512MB), 2 worker (4GB)` | `3 controller (1GB), 3 etcd (512MB), 3 worker (4GB)`
  24. fjudith revised this gist Apr 18, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -13,7 +13,7 @@ https://blog.h2o.ai/2018/03/h2o-kubeflow-kubernetes-how-to/

    Minimum | Recommandé
    ------- | ----------
    2 CPU-core 16 GB RAM | 4-core 32GB RAM
    **4 virtualcore** = 2 CPU-core 16 GB RAM + Hyperthreading | **8 Virtual core**: 4-core 32GB RAM + Hyperthreading
    Window 10 Fall Creator Update | Window 10 Fall Creator Update
    Kubernetes cluster 1.9+ [coreos-kubernetes](https://github.com/fjudith/coreos-kubernetes) |
    `1 controller (1GB), 1 etcd (512MB), 2 worker (4GB)` | `3 controller (1GB), 3 etcd (512MB), 3 worker (4GB)`
  25. fjudith revised this gist Apr 18, 2018. 1 changed file with 1 addition and 0 deletions.
    1 change: 1 addition & 0 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -17,6 +17,7 @@ Minimum | Recommandé
    Window 10 Fall Creator Update | Window 10 Fall Creator Update
    Kubernetes cluster 1.9+ [coreos-kubernetes](https://github.com/fjudith/coreos-kubernetes) |
    `1 controller (1GB), 1 etcd (512MB), 2 worker (4GB)` | `3 controller (1GB), 3 etcd (512MB), 3 worker (4GB)`
    100GB SSD (5 heures) | 360GB SSD (permanant)

    * [Chocolatey](https://chocolatey.org)
    * [ConEmu](https://conemu.github.io)
  26. fjudith revised this gist Apr 17, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -178,7 +178,7 @@ cd ~/h2oworkspace/${APPNAME}
    # add ksonnet registry to app containing all the kubeflow manifests as maintained by Google Kubeflow team
    ks registry add kubeflow https://github.com/kubeflow/kubeflow/tree/master/kubeflow
    # add ksonnet registry to app containing all the h2o component manifests
    ks registry add kubeflow ${REGISTRY}
    ks registry add h2o-kubeflow ${REGISTRY}
    ks pkg install kubeflow/core
    ks pkg install kubeflow/tf-serving
    ks pkg install kubeflow/tf-job
  27. fjudith revised this gist Apr 17, 2018. 1 changed file with 1 addition and 1 deletion.
    2 changes: 1 addition & 1 deletion h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -164,7 +164,7 @@ export GITHUB_TOKEN=
    #

    APPNAME="kubeflow-demo"
    REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/kubeflow"
    REGISTRY="github.com/fjudith/h2o-kubeflow/tree/master/h2o-kubeflow"
    NAMESPACE=kubeflow
    KF_ENV=local
    H2O3_IMAGE="quay.io/fjudith/h2o3"
  28. fjudith revised this gist Apr 17, 2018. 1 changed file with 10 additions and 4 deletions.
    14 changes: 10 additions & 4 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -175,11 +175,14 @@ pushd ~/h2oworkspace
    ks init ${APPNAME}
    cd ~/h2oworkspace/${APPNAME}

    # add ksonnet registry to app containing all the kubeflow manifests as maintained by Google Kubeflow team
    ks registry add kubeflow https://github.com/kubeflow/kubeflow/tree/master/kubeflow
    # add ksonnet registry to app containing all the h2o component manifests
    ks registry add kubeflow ${REGISTRY}
    ks pkg install kubeflow/core
    ks pkg install kubeflow/tf-serving
    ks pkg install kubeflow/tf-job
    ks pkg install kubeflow/h2o3
    ks pkg install h2o-kubeflow/h2o3-static

    kubectl create namespace ${NAMESPACE}
    ks generate core kubeflow-core --name="kubeflow-core" --namespace="${NAMESPACE}"
    @@ -263,12 +266,15 @@ kubectl apply -f local-volumes.yaml
    Exécuter les commandes suivantes pour initialiser le serveur H2O3.

    ```bash
    ks prototype use io.ksonnet.pkg.h2o3 h2o3 \
    --name h2o3 \
    ks prototype use io.ksonnet.pkg.h2o3-static h2o3-static \
    --name h2o3-static \
    --namespace kubeflow \
    --memory 1 \
    --cpu 1 \
    --replicas 2 \
    --model_server_image ${H2O3_IMAGE}

    ks apply ${KF_ENV} -c h2o3 -n kubeflow
    ks apply ${KF_ENV} -c h2o3-static -n kubeflow
    ```

    ## Déployer le Notebook H2O
  29. fjudith revised this gist Apr 15, 2018. 1 changed file with 2 additions and 2 deletions.
    4 changes: 2 additions & 2 deletions h2o-kubeflow_guide_fre.md
    View file
    Original file line number Diff line number Diff line change
    @@ -191,7 +191,7 @@ ks apply ${KF_ENV} --component="kubeflow-core"
    Valider que tous les Pods sont opérationnels

    ```bash
    kubectl get pods --namespace ${KF_ENV}
    kubectl get pods --namespace ${NAMESPACE}
    ```

    ```text
    @@ -206,7 +206,7 @@ tf-job-operator-845d5686d4-nhw6k 1/1 Running 0 21h
    Valider que les services sont bien en écoute.

    ```bash
    kubectl get svc --namespace "${KF_ENV}"
    kubectl get svc --namespace "${NAMESPACE}"
    ```

    ```text
  30. fjudith renamed this gist Apr 14, 2018. 1 changed file with 0 additions and 0 deletions.
    0 h2o-kubeflow_guide.md → h2o-kubeflow_guide_fre.md
    View file
    File renamed without changes.