Kubernetes: Getting to production

Are you ready to deploy and test Weaviate on a self-managed K8s (Kubernetes) cluster? This guide shows how to validate Weaviate’s capabilities in your enterprise environment.

At the end of this guide, expect to have:

A configured Helm-based deployment and networking setup
Basic scaling, persistent storage, and resource management
TLS, RBAC, and security best practices implemented
Monitoring, logging, and backup strategies enabled

Prerequisites

Before beginning, ensure that you have the following:

Basic Kubernetes and containerization conceptual knowledge
Basic experience with Helm and kubectl
A running Kubernetes cluster with Weaviate installed
kubectl installed
Helm installed

Step 1: Configure your Helm Chart

Use the official Weaviate Helm chart for your installation:

  helm repo add weaviate https://weaviate.github.io/weaviate-helm
  helm install my-weaviate weaviate/weaviate

Customize the values to fit your enterprise requirements (e.g., resource allocation, storage settings).
Deploy the chart and verify pod health.

Step 2: Network Security

Configure an ingress controller to securely expose Weaviate.
Enable TLS with a certificate manager and enforce TLS encryption for all client-server communication.
Assign a domain name for external access.
Implement RBAC or admin lists to restrict user access.

An example of RBAC enabled on your Helm chart

YAML

authorization:
  rbac:
    enabled: true
    root_users:
      - admin_user1
      - admin_user2

An example of admin lists implemented on your Helm chart (if not using RBAC)

YAML

admin_list:
  enabled: true
  users:
    - admin_user1
    - admin_user2
    - api-key-user-admin
  read_only_users:
    - readonly_user1
    - readonly_user2
    - api-key-user-readOnly

Admin List Configuration

Step 3: Scaling

Implement horizontal scaling to ensure high availability:

YAML

replicaCount: 3

Define CPU/memory limits and requests to optimize pod efficiency.

An example of defining CPU and memory limits and cores

YAML

resources:
  requests:
    cpu: "500m"
    memory: "1Gi"
  limits:
    cpu: "2"
    memory: "4Gi"

Step 4: Monitoring and Logging

Use Prometheus and Grafana to collect and analyze performance metrics.
Implement alerting for issue resolution.

An example of enabling service monitoring

YAML

serviceMonitor:
  enabled: true
  interval: 30s
  scrapeTimeout: 10s

Step 5: Upgrades and Backups

Use the rolling update strategy used by Helm to minimize downtime.

An example of configuring the rolling update strategy.

YAML

updateStrategy:
  type: RollingUpdate
  rollingUpdate:
    maxSurge: 1
    maxUnavailable: 0

Test new Weaviate versions before deploying into production.
Implement disaster recovery procedures to ensure that data is restored quickly.

Conclusion

Voila! You now have a deployment that is somewhat ready for production. Your next step will be to complete the self-assessment and identify any gaps.

Next Steps: Production Readiness Self-Assessment

Questions and feedback

Have a question or feedback? Here's how to reach us.

Community Forum

Ask questions and connect with other developers on our Community forum.

Support

Weaviate Cloud user or customer? Find the right channel on the Support page.