構成

dism /online /Enable-Feature /FeatureName:ServicesForNFS-ClientOnly
dism /online /Enable-Feature /FeatureName:ClientForNFS-Infrastructure

New-ItemProperty -Path "HKLM:\Software\Microsoft\ClientForNFS\CurrentVersion\Default" -Name "AnonymousGid" -Value 65532 -PropertyType DWord
New-ItemProperty -Path "HKLM:\Software\Microsoft\ClientForNFS\CurrentVersion\Default" -Name "AnonymousUid" -Value 65532 -PropertyType DWord

nfsadmin client stop
nfsadmin client start

kubectl port-forward -n abbyy-infrastructure service/consul-ui 8500:80

label=mail
kubectl -n abbyy-vantage rollout restart $(kubectl -n abbyy-vantage get deployments -l app.kubernetes.io/component=$label -o name)

apt-get update
apt-get install dkms
dpkg -i nvidia-linux-grid-535_535.54.03_amd64.deb

docker run -it \
-v $PWD/kube:/root/.kube \
-v $PWD/ssh/ansible:/root/.ssh/ansible \
-v "//var/run/docker.sock:/var/run/docker.sock" \
-v $PWD/inventory:/ansible/inventories/k8s/inventory \
-v $PWD/env_specific.yml:/ansible/inventories/k8s/group_vars/all/env_specific.yml \
-v $PWD/ssl:/ansible/files/ssl:ro \
-v $PWD/gpu:/ansible/files/gpu:ro \
--privileged \
registry.local/vantage/vantage-k8s:2.7.1

[abbyy_workers]
worker16-48-w01 ansible_host=10.10.10.27
worker16-48-w02 ansible_host=10.10.10.21
worker16-48-w03 ansible_host=10.10.10.20
worker2-12-a40-gpu01 ansible_host=10.10.10.60

chmod 600 /root/.ssh/ansible
ansible-playbook -i inventories/k8s -v playbooks/4-Kubernetes-k8s.yml

ansible-playbook -i inventories/k8s -v playbooks/setup-gpu-node.yml

docker run -it \
-v $PWD/kube:/root/.kube \
-v $PWD/ssh/ansible:/root/.ssh/ansible \
-v "//var/run/docker.sock:/var/run/docker.sock" \
-v $PWD/inventory:/ansible/inventories/k8s/inventory \
-v $PWD/env_specific.yml:/ansible/inventories/k8s/group_vars/all/env_specific.yml \
-v $PWD/ssl:/ansible/files/ssl:ro \
--privileged \
registry.local/vantage/vantage-k8s:2.7.1

[abbyy_workers]
worker16-48-w01 ansible_host=10.10.10.27
worker16-48-w02 ansible_host=10.10.10.21
worker16-48-w03 ansible_host=10.10.10.20
worker2-12-a40-gpu01 ansible_host=10.10.10.60

ansible-playbook -i inventories/k8s -v playbooks/4-Kubernetes-k8s.yml

helm upgrade --install gpu-operator ansible/files/helm/charts/gpu-operator --create-namespace --debug -n gpu-operator

kubectl taint nodes worker2-12-a40-gpu01 nvidia.com/gpu:NoSchedule

kubectl apply -f filename

apiVersion: v1
kind: Pod
metadata:
  name: gpu-pod
  namespace: gpu-operator
spec:
  restartPolicy: Never
  containers:
    - name: cuda-container
      image: nvcr.io/nvidia/k8s/cuda-sample:vectoradd-cuda10.2
      resources:
        limits:
          nvidia.com/gpu: 1 # 1つのGPUをリクエスト
  tolerations:
    - key: nvidia.com/gpu
      operator: Exists
      effect: NoSchedule
    - key: k8s.abbyy.com/techcore
      effect: NoSchedule
      value: "true"

kubectl -n gpu-operator logs gpu-pod

[Vector addition of 50000 elements]
Copy input data from the host memory to the CUDA device
CUDA kernel launch with 196 blocks of 256 threads
Copy output data from the CUDA device to the host memory
Test PASSED
Done

techcore:
  use_gpu_workers: true
  use_nn_extraction_training_workers: true

ansible-playbook -i inventories/k8s -v playbooks/11-DeployWorkers-k8s.yml

kubectl port-forward -n abbyy-infrastructure service/consul-ui 8500:80

kubectl -n abbyy-vantage rollout restart $(kubectl -n abbyy-vantage get deployments -l app.kubernetes.io/component=verification -o name)

kubectl -n abbyy-vantage rollout restart $(kubectl -n abbyy-vantage get deployments -l app.kubernetes.io/component=manualverification -o name)

kubectl get nodes

-----BEGIN CERTIFICATE-----
[your certificate]
-----END CERTIFICATE-----

-----BEGIN PRIVATE KEY-----
[your key]
-----END PRIVATE KEY-----

for i in `kubectl get secret --field-selector metadata.name=platform-wildcard -o custom-columns=:metadata.namespace -A --no-headers 2>/dev/null`; do kubectl patch secret platform-wildcard -p "{\"data\":{\"tls.key\":\"$(base64 < "./key.pem" | tr -d '\n')\", \"tls.crt\":\"$(base64 < "./cert.pem" | tr -d '\n')\"}}" -n $i; done

kubectl rollout restart deployment -n abbyy-infrastructure $(kubectl get deployment -n abbyy-infrastructure -o custom-columns=NAME:metadata.name --no-headers 2>/dev/null | grep ingress-nginx-controller)