This integration is available for Portkey Enterprise users.
When deploying your AI app to production, having a robust, scalable, and high-performance logging solution is crucial. That’s where Portkey and MongoDB combine — Portkey’s Enterprise version easily lets you store all your LLM logs on the most popular database services of your choice. This is made even easier with our ready-to-use Kubernetes Configs (Helm charts).
Portkey is part of MongoDB partner ecosystem to help you build & deploy your AI apps with confidence. Learn more
Getting Started with MongoDB Log Storage
Prerequisites
- Portkey Enterprise account
- Access to a MongoDB instance
- Kubernetes cluster
Configuration
To use MongoDB for Log storage, you’ll need to provide the following values in your values.yaml file for the Helm chart deployment:
MONGO_DB_CONNECTION_URL:
MONGO_DATABASE:
MONGO_COLLECTION_NAME:
MONGO_GENERATION_HOOKS_COLLECTION_NAME:
Authentication with PEM File
If you’re using a PEM file for authentication, follow these additional steps:
Add PEM File Details
In the resources-config.yaml file, supply PEM file details under the data section:data:
document_db.pem: |
-----BEGIN CERTIFICATE-----
Your certificate content here
-----END CERTIFICATE-----
Modify value.yaml
In values.yaml, add the following configuration:volumes:
- name: shared-folder
configMap:
name: resource-config
volumeMounts:
- name: shared-folder
mountPath: /etc/shared/<shared_pem>
subPath: <shared_pem>
Use the PEM File
Update your MONGO_DB_CONNECTION_URL to use the PEM file:mongodb://<user>:<password>@<host>?tls=true&tlsCAFile=/etc/shared/document_db.pem&retryWrites=false
Cloud Deployment
Portkey with MongoDB integration can be deployed to all major cloud providers. For cloud-specific documentation, please refer to:
Get started with Portkey Enterprise here.