ops: Creating Chart. Added postgres

Author: alhendrickson

deployment/kubernetes/charts/medcat-trainer-helm/Chart.lock

@@ -2,5 +2,8 @@ dependencies:
 - name: solr
   repository: oci://registry-1.docker.io/bitnamicharts
   version: 9.6.10
-digest: sha256:f62be9fbd53de5aebf148b21c3ce7ff568a4d196f77f99a14c1207755d55e0e5
-generated: "2025-09-08T10:31:52.835071366Z"
+- name: postgresql
+  repository: oci://registry-1.docker.io/bitnamicharts
+  version: 16.7.27
+digest: sha256:a02db326b15b24d92e4c0787792803d7b224babd4f288e28a73d66a7d4506a70
+generated: "2025-09-10T11:29:31.705470067Z"

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

@@ -28,3 +28,7 @@ dependencies:
 - name: solr
   version: "9.6.10"
   repository: "oci://registry-1.docker.io/bitnamicharts"
+- name: postgresql
+  version: 16.7.27
+  repository: "oci://registry-1.docker.io/bitnamicharts"
+  condition: postgresql.enabled

deployment/kubernetes/charts/medcat-trainer-helm/README.md

@@ -0,0 +1,45 @@
+# MedCAT Trainer Helm Chart
+
+This Helm chart deploys MedCAT Trainer and infrastructure to a Kubernetes cluster.
+
+By default the chart will:
+
+- Run MedCAT Trainer Django server
+- Run NGINX for static site hosting and routing
+- Run a SOLR and Zookeeper cluster for the Concept DB
+- Run a Postgres database for persistence
+
+
+## Installation
+
+```sh
+helm install my-medcat-trainer oci://registry-1.docker.io/cogstacksystems/medcat-trainer-helm
+```
+
+## Configuration
+
+See these values for common configurations to change:
+
+| Setting  |description    |
+| -------- | -------- |
+| `env`          | Environment variables as defined in the [MedCAT Trainer docs](https://docs.cogstack.org/projects/medcat-trainer/en/latest/installation.html).    |
+|`medcatConfig`|MedCAT config file as described [here](https://github.com/CogStack/cogstack-nlp/blob/main/medcat-v2/medcat/config/config.py)|
+
+### Use Sqlite instead of Postgres
+
+Sqlite can be used for smaller single instance deployments
+
+Set these values:
+
+```yaml
+DB_ENGINE: "sqlite3"
+
+postgresql:
+  enabled: false
+```
+
+### Missing features
+These features are not yet existing but to be added in future:
+- Use a pre existing postgres db
+- Use a pre existing SOLR instnace
+- Migrate from supervisord to standalone pods for background tasks for scaling purposes

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

@@ -54,6 +54,16 @@ spec:
           envFrom:
             - configMapRef:
                 name: {{ include "medcat-trainer-helm.fullname" . }}-medcat-trainer-env
+          {{- if .Values.postgresql.enabled }}
+          env:
+            - name: DB_HOST
+              value: {{ .Release.Name }}-postgresql
+            - name: DB_PASSWORD
+              valueFrom:
+                secretKeyRef:
+                  name: {{ .Release.Name }}-postgresql
+                  key: postgres-password
+          {{- end }}
           {{- with .Values.livenessProbe }}
           livenessProbe:
             {{- toYaml . | nindent 12 }}

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

@@ -171,7 +171,7 @@ data:
             proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
         }
 
-        location /healthz {
+        location /nginx/health/live {
           access_log off;
           return 200 'OK';
           add_header Content-Type text/plain;

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

@@ -7,7 +7,7 @@ metadata:
     app.kubernetes.io/component: nginx
 spec:
   {{- if not .Values.autoscaling.enabled }}
-  replicas: {{ .Values.replicaCount }}
+  replicas: {{ .Values.nginxReplicaCount }}
   {{- end }}
   selector:
     matchLabels:

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

@@ -32,7 +32,7 @@ spec:
     requests:
       storage: {{.Values.persistence.static.size}}
 ---
-#TODO - only create if type using is sqlite
+{{- if eq .Values.DB_ENGINE "sqlite3" }}
 apiVersion: v1
 kind: PersistentVolumeClaim
 metadata:
@@ -49,8 +49,9 @@ spec:
   resources:
     requests:
       storage: {{.Values.persistence.sqlite.size}}
+{{- end }}
 ---
-#TODO - only create if type using is sqlite
+{{- if eq .Values.DB_ENGINE "sqlite3" }}
 apiVersion: v1
 kind: PersistentVolumeClaim
 metadata:
@@ -66,4 +67,5 @@ spec:
     - ReadWriteOnce
   resources:
     requests:
-      storage: {{.Values.persistence.sqlite.backupDbSize}}
+      storage: {{.Values.persistence.sqlite.backupDbSize}}
+{{- end }}

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

@@ -4,6 +4,7 @@
 
 # This will set the replicaset count more information can be found here: https://kubernetes.io/docs/concepts/workloads/controllers/replicaset/
 replicaCount: 1
+nginxReplicaCount: 3
 
 # This sets the container image more information can be found here: https://kubernetes.io/docs/concepts/containers/images/
 image:
@@ -20,7 +21,6 @@ nginxImage:
 # Add any environment variables here that should be set in the medcat-trainer container
 env:
   CSRF_TRUSTED_ORIGINS: "http://localhost:8000"
-  DB_ENGINE: "sqlite3"
   DEBUG: "1"
   EMAIL_HOST: "mail.cogstack.org"
   EMAIL_PASS: "to-be-changed"
@@ -35,7 +35,27 @@ env:
   RESUBMIT_ALL_ON_STARTUP: "0"
   UNIQUE_DOC_NAMES_IN_DATASETS: "True"
 
-
+  # TODO: Support custom DB overrides#
+  # DB_ENGINE: "sqlite3"
+  DB_ENGINE: "postgresql"
+  DB_NAME: "postgres"
+  DB_USER: "postgres"
+  DB_PORT: "5432"
+  # DB_PASSWORD: ""
+  # DB_HOST: "" 
+
+postgresql:
+  enabled: true
+  # TODO: Support custom DB overrides
+  # auth:
+  #  - name for a custom database
+  #   database: "my_trainer_db"
+  #   username: "trainer_admin"
+  #   password: "changeme_changeme"
+  primary:
+    persistence:
+      # Size of the PVC for the postgres database
+      size: 500Mi
 persistence:
   media:
     # Size of PVC for files like model packs and other media downloaded by medcat trainer
@@ -49,8 +69,7 @@ persistence:
     # Size of the PVC for the Sqlite backups
     backupDbSize: 300Mi
   storageClassName: ""
-  # Engine can be 'sqlite3' or 'postgresql'
-  dbEngine: sqlite3
+
 
 # MedCAT config as described here: https://github.com/CogStack/cogstack-nlp/blob/main/medcat-v2/medcat/config/config.py
 medcatConfig: |
@@ -160,22 +179,30 @@ resources: {}
 
 # 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/
 # TODO Add liveness and readiness to django app
+# https://github.com/CogStack/cogstack-nlp/pull/131/files
 # livenessProbe:
 #   httpGet:
-#     path: /
+#     path: /api/health/live/
 #     port: http
 # readinessProbe:
 #   httpGet:
-#     path: /
+#     path: /api/health/ready/
 #     port: http
+# startupProbe:
+#   httpGet:
+#     path: /api/health/startup/
+#     port: http
+#   failureThreshold: 30
+#   periodSeconds: 10
+
 nginx:
   livenessProbe:
     httpGet:
-      path: /healthz
+      path: /nginx/health/live
       port: http
   readinessProbe:
     httpGet:
-      path: /healthz
+      path: /nginx/health/live
       port: http
 
 

deployment/kubernetes/local_dev_startup.sh

@@ -21,5 +21,5 @@ helm test medcat-service --logs
 # Test medcat trainer
 # kubectl port-forward svc/nginx 8000:8000
 
-helm upgrade medcat-trainer ./medcat-trainer-helm --install --recreate-pods  --wait --timeout 5m0s # Install if it doesnt already exist, else upgrade
+helm upgrade my-test ./medcat-trainer-helm --install --recreate-pods  --wait --timeout 5m0s # Install if it doesnt already exist, else upgrade
 # kubectl port-forward svc/medcat-trainer-solr 8983:8983