feat(helm): In helm add update strategy value for deployments. (#27)

* feat(helm): Add example env vars for single threading * build: Only run github actions based on folder changes * build: Update helm github action

Author: alhendrickson

.github/workflows/kubernetes-charts-build.yaml

@@ -1,11 +1,19 @@
 name: Kubernetes - Lint, Test, and Publish Helm Charts
 
-on: pull_request
+on:
+ push:
+ branches: [ main ]
+ pull_request:
+ paths:
+ - 'deployment/kubernetes/**'
+ - '.github/workflows/kubernetes**'
+ tags:
+ - 'v*.*.*' # e.g., v0.1.1
+
 defaults:
  run:
  working-directory: ./deployment/kubernetes
 
-
 jobs:
  helm-lint-test:
  runs-on: ubuntu-latest
@@ -17,9 +25,9 @@ jobs:
  fetch-depth: 0
 
  - name: Set up Helm
- uses: azure/setup-helm@v4.2.0
+ uses: azure/setup-helm@v4.3.0
  with:
- version: v3.17.0
+ version: v3.18.3
 
  - uses: actions/setup-python@v5.3.0
  with:

.github/workflows/observability-docker.yml

@@ -2,8 +2,11 @@ name: ci-build
 
 on:
  push:
- branches:
- - '**'
+ branches: [ main ]
+ pull_request:
+ paths:
+ - 'observability/**'
+ - '.github/workflows/observability**'
  tags:
  - 'v*.*.*' # e.g., v0.1.1
 

deployment/kubernetes/charts/medcat-service-helm/templates/deployment.yaml

@@ -11,6 +11,7 @@ spec:
  selector:
  matchLabels:
  {{- include "medcat-service.selectorLabels" . | nindent 6 }}
+ strategy: {{ toYaml .Values.updateStrategy | nindent 4 }}
  template:
  metadata:
  {{- with .Values.podAnnotations }}

deployment/kubernetes/charts/medcat-service-helm/values.yaml

@@ -31,6 +31,19 @@ env:
  # DEID_MODE: true
  # DEID_REDACT: true
 
+ # Set SERVER_GUNICORN_MAX_REQUESTS to a high number instead of the default 1000. Trust k8s instead to restart pod when needed.
+ SERVER_GUNICORN_MAX_REQUESTS: 1000000
+
+ # Recommended env vars to set to try to limit to 1 CPU for scaling
+ # OMP_NUM_THREADS: "1"
+ # OPENBLAS_NUM_THREADS: "1"
+ # MKL_NUM_THREADS: "1"
+ # VECLIB_MAXIMUM_THREADS: "1"
+ # NUMEXPR_NUM_THREADS: "1"
+ # TOKENIZERS_PARALLELISM: "false"
+ # PYTORCH_ENABLE_MPS_FALLBACK: "1"
+ # SERVER_GUNICORN_EXTRA_ARGS: "--worker-connections 1 --backlog 1"
+
 # Enable downloading of public models using wget on startup. Model will be downloaded to /models/<name> and used for APP_MEDCAT_MODEL_PACK
 model: {}
  # Public URL to download a model pack from
@@ -44,6 +57,14 @@ imagePullSecrets: []
 nameOverride: ""
 fullnameOverride: ""
 
+updateStrategy:
+ # Used for Kubernetes deployment .spec.strategy.type. Allowed values are "Recreate" or "RollingUpdate".
+ type: RollingUpdate
+ # RollingUpdate default options example
+ # rollingUpdate:
+ # maxSurge: 25%
+ # maxUnavailable: 25%
+
 # This section builds out the service account more information can be found here: https://kubernetes.io/docs/concepts/security/service-accounts/
 serviceAccount:
  # Specifies whether a service account should be created
@@ -108,11 +129,11 @@ resources: {}
  # resources, such as Minikube. If you do want to specify resources, uncomment the following
  # lines, adjust them as necessary, and remove the curly braces after 'resources:'.
  # limits:
- # cpu: 100m
- # memory: 128Mi
+ # cpu: 2 # Note one general k8s recommendation is to leave CPU limit unset
+ # memory: 4Gi
  # requests:
- # cpu: 100m
- # memory: 128Mi
+ # cpu: 1
+ # memory: 2Gi
 
 # This is to setup the liveness and readiness probes more information can be found here: https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
 livenessProbe:
@@ -129,6 +150,7 @@ startupProbe:
  port: http
  failureThreshold: 30
  periodSeconds: 10
+ initialDelaySeconds: 2
 
 # This section is for setting up autoscaling more information can be found here: https://kubernetes.io/docs/concepts/workloads/autoscaling/
 autoscaling:

deployment/kubernetes/charts/medcat-trainer-helm/templates/medcat-trainer-deployment.yaml

@@ -13,6 +13,7 @@ spec:
  matchLabels:
  {{- include "medcat-trainer-helm.selectorLabels" . | nindent 6 }}
  app.kubernetes.io/component: medcat-trainer
+ strategy: {{ toYaml .Values.updateStrategy | nindent 4 }}
  template:
  metadata:
  {{- with .Values.podAnnotations }}

deployment/kubernetes/charts/medcat-trainer-helm/templates/nginx-deployment.yaml

@@ -13,6 +13,7 @@ spec:
  matchLabels:
  {{- include "medcat-trainer-helm.selectorLabels" . | nindent 6 }}
  app.kubernetes.io/component: nginx
+ strategy: {{ toYaml .Values.nginxUpdateStrategy | nindent 4 }}
  template:
  metadata:
  {{- with .Values.podAnnotations }}

deployment/kubernetes/charts/medcat-trainer-helm/values.yaml

@@ -111,6 +111,14 @@ imagePullSecrets: []
 nameOverride: ""
 fullnameOverride: ""
 
+updateStrategy:
+ # Used for Kubernetes deployment .spec.strategy.type. Allowed values are "Recreate" or "RollingUpdate".
+ type: RollingUpdate
+
+nginxUpdateStrategy:
+ # Used for Kubernetes deployment .spec.strategy.type. Allowed values are "Recreate" or "RollingUpdate".
+ type: RollingUpdate
+
 # This section builds out the service account more information can be found here: https://kubernetes.io/docs/concepts/security/service-accounts/
 serviceAccount:
  # Specifies whether a service account should be created