Added templates to deploy in-cluster minio and milvus

.gitignore

@@ -1 +1,2 @@
-portkey_config*.yaml
+portkey_config*.yaml
+**/.DS_Store

charts/portkey-gateway/README.md

@@ -42,6 +42,9 @@ Choose your storage backends from the options below. You'll need to configure:
 **Cache Store**
 - Cache configuration
 
+**Vector Store** *(Optional)*
+- For semantic caching with Milvus. See [Vector Store Setup](./docs/VectorStore.md)
+
 ---
 
 ## Analytics Store
@@ -241,6 +244,40 @@ LOG_STORE_BASEPATH: "<Custom S3 Base Path Including Bucket Name>"
 ```
 </details>
 
+### MinIO (Local/In-Cluster)
+Simple deployments, development
+
+```yaml
+
+# Local (MinIO) object store configuration section
+environment:
+  data: 
+    LOG_STORE: local
+    LOG_STORE_GENERATIONS_BUCKET: "<Bucket Name>"                 # Bucket will automatically be create in MinIO by this Helm chart. Default portkey-log-store
+
+minio:
+  name: "minio"
+  authKey:
+    create: true
+    accessKey: "portkey"
+    secretKey: "portkey123"
+  persistence:
+    enabled: true
+    size: 10Gi
+    storageClassName: "<Storage Class Name>"                      # Provide k8s storage class name to provision volumes for data persistence.
+    accessMode: ReadWriteOnce   
+```
+
+**Notes**
+
+* The values specified for `minio.authKey.accessKey` and `minio.authKey.secretKey` will be used as the MinIO `ROOT USERNAME` and `ROOT PASSWORD`.
+* To access the MinIO WebUI console, port-forward the service to your local machine:
+  ```sh
+  kubectl port-forward svc/minio 9001:9001
+  ```
+  Then open `http://localhost:9001` in your browser.
+
+
 ### Azure Blob Storage
 <details>
 <summary>Detailed Azure Blob Storage Setup</summary>
@@ -451,4 +488,4 @@ helm uninstall portkey-gateway --namespace portkeyai
 For additional help:
 - Check the [full configuration reference](values.yaml)
 - Review logs: `kubectl logs -n portkeyai deployment/portkey-gateway`
-- Contact support(suport@portkey.ai) with your configuration details
+- Contact support(support@portkey.ai) with your configuration details

charts/portkey-gateway/docs/Configuration.md

@@ -22,6 +22,15 @@
 | `images.redisImage.repository` | string | `"docker.io/redis"` | Redis container image repository |
 | `images.redisImage.pullPolicy` | string | `"IfNotPresent"` | Image pull policy for Redis |
 | `images.redisImage.tag` | string | `"7.2-alpine"` | Redis image tag |
+| `images.minioImage.repository` | string | `"minio/minio"` | MinIO container image repository |
+| `images.minioImage.pullPolicy` | string | `"IfNotPresent"` | Image pull policy for MinIO |
+| `images.minioImage.tag` | string | `"RELEASE.2025-09-07T16-13-09Z"` | MinIO image tag |
+| `images.etcdImage.repository` | string | `"quay.io/coreos/etcd"` | etcd container image repository |
+| `images.etcdImage.pullPolicy` | string | `"IfNotPresent"` | Image pull policy for etcd |
+| `images.etcdImage.tag` | string | `"v3.5.5"` | etcd image tag |
+| `images.milvusImage.repository` | string | `"milvusdb/milvus"` | Milvus container image repository |
+| `images.milvusImage.pullPolicy` | string | `"IfNotPresent"` | Image pull policy for Milvus |
+| `images.milvusImage.tag` | string | `"v2.3.21"` | Milvus image tag |
 | `imagePullSecrets` | array | `[portkeyenterpriseregistrycredentials]` | Kubernetes secrets for pulling private images |
 
 ### Image Credentials
@@ -382,4 +391,105 @@ environment:
 |----------|------|---------|-------------|
 | `redis.name` | string | `"redis"` | Redis component name |
 | `redis.containerPort` | integer | `6379` | Redis container port |
-| `redis.resources` | object | `{}` | Resource requests and limits for Redis |
+| `redis.resources` | object | `{}` | Resource requests and limits for Redis |
+
+### MinIO Configuration
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `minio.name` | string | `"minio"` | MinIO component name |
+| `minio.apiPort` | integer | `9000` | MinIO API port |
+| `minio.consolePort` | integer | `9001` | MinIO WebUI console port |
+
+#### MinIO Auth Key Secret
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `minio.authKey.create` | boolean | `true` | Create MinIO credentials secret |
+| `minio.authKey.existingSecret` | string | `""` | Use existing secret for MinIO credentials |
+| `minio.authKey.accessKey` | string | `"portkey"` | MinIO root username / access key |
+| `minio.authKey.secretKey` | string | `"portkey123"` | MinIO root password / secret key |
+
+#### MinIO StatefulSet
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `minio.statefulSet.labels` | object | `{}` | Labels for MinIO StatefulSet |
+| `minio.statefulSet.annotations` | object | `{}` | Annotations for MinIO StatefulSet |
+| `minio.statefulSet.podSecurityContext` | object | `{}` | Pod security context |
+| `minio.statefulSet.securityContext` | object | `{}` | Container security context |
+| `minio.statefulSet.resources` | object | `{}` | Resource requests and limits |
+| `minio.statefulSet.nodeSelector` | object | `{}` | Node selector for scheduling |
+| `minio.statefulSet.tolerations` | array | `[]` | Tolerations for scheduling |
+| `minio.statefulSet.affinity` | object | `{}` | Affinity rules for scheduling |
+
+#### MinIO Service
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `minio.service.type` | string | `"ClusterIP"` | MinIO service type |
+| `minio.service.labels` | object | `{}` | Service labels |
+| `minio.service.annotations` | object | `{}` | Service annotations |
+
+#### MinIO Persistence
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `minio.persistence.enabled` | boolean | `true` | Enable persistent storage for MinIO |
+| `minio.persistence.size` | string | `"10Gi"` | Storage size for MinIO data |
+| `minio.persistence.storageClassName` | string | `""` | Storage class for MinIO PVC |
+| `minio.persistence.accessMode` | string | `"ReadWriteOnce"` | PVC access mode |
+
+### Milvus Configuration
+
+> **Note**: Milvus requires MinIO for object storage. MinIO is deployed automatically when `environment.data.LOG_STORE: local` or `environment.data.VECTOR_STORE: local` is set.
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `environment.data.VECTOR_STORE` | string | `""` | Set to `local` to deploy in-cluster Milvus and auto-configure Gateway env vars |
+
+#### Milvus etcd Configuration
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `milvus.etcd.port` | integer | `2379` | etcd client port |
+| `milvus.etcd.env.autoCompactionMode` | string | `"revision"` | etcd auto compaction mode |
+| `milvus.etcd.env.autoCompactionRetention` | string | `"1000"` | etcd auto compaction retention |
+| `milvus.etcd.env.quotaBackendBytes` | string | `"4294967296"` | etcd backend quota in bytes |
+| `milvus.etcd.env.snapshotCount` | string | `"50000"` | etcd snapshot count |
+| `milvus.etcd.persistence.enabled` | boolean | `true` | Enable persistent storage for etcd |
+| `milvus.etcd.persistence.size` | string | `"10Gi"` | Storage size for etcd data |
+| `milvus.etcd.persistence.storageClassName` | string | `""` | Storage class for etcd PVC |
+| `milvus.etcd.service.serviceType` | string | `"ClusterIP"` | etcd service type |
+| `milvus.etcd.service.labels` | object | `{}` | etcd service labels |
+| `milvus.etcd.service.annotations` | object | `{}` | etcd service annotations |
+| `milvus.etcd.resources` | object | `{}` | Resource requests and limits for etcd |
+| `milvus.etcd.statefulSet.labels` | object | `{}` | Labels for etcd StatefulSet |
+| `milvus.etcd.statefulSet.annotations` | object | `{}` | Annotations for etcd StatefulSet |
+| `milvus.etcd.statefulSet.podSecurityContext` | object | `{}` | Pod security context |
+| `milvus.etcd.statefulSet.securityContext` | object | `{}` | Container security context |
+| `milvus.etcd.statefulSet.nodeSelector` | object | `{}` | Node selector for scheduling |
+| `milvus.etcd.statefulSet.tolerations` | array | `[]` | Tolerations for scheduling |
+| `milvus.etcd.statefulSet.affinity` | object | `{}` | Affinity rules for scheduling |
+
+#### Milvus Server Configuration
+
+| Variable | Type | Default | Description |
+|----------|------|---------|-------------|
+| `milvus.grpcPort` | integer | `19530` | Milvus gRPC port |
+| `milvus.httpPort` | integer | `9091` | Milvus HTTP port |
+| `milvus.persistence.enabled` | boolean | `true` | Enable persistent storage for Milvus |
+| `milvus.persistence.size` | string | `"10Gi"` | Storage size for Milvus data |
+| `milvus.persistence.storageClassName` | string | `""` | Storage class for Milvus PVC |
+| `milvus.persistence.accessMode` | string | `"ReadWriteOnce"` | PVC access mode |
+| `milvus.service.serviceType` | string | `"ClusterIP"` | Milvus service type |
+| `milvus.service.labels` | object | `{}` | Milvus service labels |
+| `milvus.service.annotations` | object | `{}` | Milvus service annotations |
+| `milvus.resources` | object | `{}` | Resource requests and limits for Milvus |
+| `milvus.statefulSet.labels` | object | `{}` | Labels for Milvus StatefulSet |
+| `milvus.statefulSet.annotations` | object | `{}` | Annotations for Milvus StatefulSet |
+| `milvus.statefulSet.podSecurityContext` | object | `{}` | Pod security context |
+| `milvus.statefulSet.securityContext` | object | `{}` | Container security context |
+| `milvus.statefulSet.nodeSelector` | object | `{}` | Node selector for scheduling |
+| `milvus.statefulSet.tolerations` | array | `[]` | Tolerations for scheduling |
+| `milvus.statefulSet.affinity` | object | `{}` | Affinity rules for scheduling |

charts/portkey-gateway/docs/VectorStore.md

@@ -0,0 +1,115 @@
+# Vector Store Configuration
+
+This guide explains how to configure local/in-cluster Milvus as a vector store for semantic caching in Portkey Gateway.
+
+## Enable MinIO in Helm Chart
+
+Milvus requires MinIO for object storage. Ensure MinIO is enabled in your deployment before proceeding. See [MinIO setup in README](../README.md#minio-localin-cluster) for configuration details.
+
+**Note:** If local MinIO is already enabled for log storage, skip this section and proceed to the next step.
+
+## Enable Milvus in Helm Chart
+
+Add the following to your `values.yaml`:
+
+```yaml
+environment:
+  data: 
+    VECTOR_STORE: local
+
+milvus:
+  name: "milvus"
+  persistence:
+    enabled: true
+    size: 10Gi
+    storageClassName: ""                                # Provide k8s storage class for creating storage volumes. 
+    accessMode: ReadWriteOnce
+  etcd:
+    name: "etcd"
+    persistence:
+      enabled: true
+      size: 10Gi
+      storageClassName: "gp3"                           # Provide k8s storage class for creating storage volumes. 
+      accessMode: ReadWriteOnce  
+```
+
+## Access Milvus Instance
+
+After deploying, port-forward the Milvus service to access it locally:
+
+```sh
+kubectl port-forward -n portkeyai svc/milvus 19530:19530
+```
+
+## Connect to Milvus
+
+You can use a client like [Attu](https://github.com/zilliztech/attu) to connect to the Milvus instance:
+
+| Field | Value |
+|-------|-------|
+| Host | `localhost:19530` |
+| Username | `root` |
+| Password | `Milvus` |
+
+## Create Collection
+
+Create the required collection for semantic caching. The example below creates a collection named `textEmbedding3Small` with 1536 dimensions (update `dim` based on your embedding model):
+
+```sh
+curl --location --request POST 'http://localhost:19530/v2/vectordb/collections/create' \
+--header 'Authorization: Bearer root:Milvus' \
+--header 'Content-Type: application/json' \
+--data '{
+  "collectionName": "textEmbedding3Small",
+  "schema": {
+    "autoId": false,
+    "enableDynamicField": true,
+    "fields": [
+      {
+        "fieldName": "id",
+        "dataType": "VarChar",
+        "isPrimary": true,
+        "elementTypeParams": {
+          "max_length": 1024
+        }
+      },
+      {
+        "fieldName": "vector",
+        "dataType": "FloatVector",
+        "elementTypeParams": {
+          "dim": "1536"
+        }
+      },
+      {
+        "fieldName": "workspace_id",
+        "dataType": "VarChar",
+        "elementTypeParams": {
+          "max_length": 512
+        }
+      }
+    ]
+  },
+  "indexParams": [
+    {
+      "fieldName": "vector",
+      "metricType": "COSINE",
+      "indexName": "vector",
+      "indexType": "AUTOINDEX"
+    }
+  ],
+  "enableDynamicField": true
+}'
+```
+
+### Verify Collection
+
+Confirm the collection was created successfully:
+
+```sh
+curl --location --request POST 'http://localhost:19530/v2/vectordb/collections/describe' \
+--header 'Authorization: Bearer root:Milvus' \
+--header 'Content-Type: application/json' \
+--data '{
+    "collectionName": "textEmbedding3Small"
+}'
+```