Weaviate Helm Chart AKS Internal Load Balancer Example

Published
Length
3 min read

Luis Felipe Dias

All posts

A weaviate helm chart aks internal load balancer example deploys your vector database on a private Azure network. You configure the Helm service block to request an internal cloud load balancer instead of a public IP. Teams use this pattern so internal apps can query vector data without exposing endpoints to the internet. Still, private network routing alone does not restrict what engineers or workloads do once inside the cluster. When teams grant broad Kubernetes roles or share database keys,

Your coding agents run with your access.

Claude Code reaches prod with a developer's credentials, and approval prompts get clicked through. hoop.dev checks every query, masks sensitive data, and holds risky changes for a human.

A weaviate helm chart aks internal load balancer example deploys your vector database on a private Azure network. You configure the Helm service block to request an internal cloud load balancer instead of a public IP.

Teams use this pattern so internal apps can query vector data without exposing endpoints to the internet. Still, private network routing alone does not restrict what engineers or workloads do once inside the cluster. When teams grant broad Kubernetes roles or share database keys, any user on the network can read or change sensitive vectors.

How a weaviate helm chart aks internal load balancer example works

When you deploy Weaviate on Azure Kubernetes Service with Helm, the chart creates a Kubernetes service resource. By default, a load balancer service asks Azure for a public IP address. To keep traffic private, you add an Azure internal load balancer annotation to the service settings in your Helm values.

Azure then assigns a private IP from your virtual network subnet to the Weaviate service. Only pods, virtual machines, and peered networks inside that private boundary can reach the database port. You also configure the cluster role-based rules to control who can change this deployment.

In Kubernetes, a Role sets permissions within one namespace, while a ClusterRole can cover all namespaces. Permissions are additive, and there are no deny rules in the native model. A RoleBinding grants those namespace permissions to a user or group, while a ClusterRoleBinding grants access across the whole cluster.

How Kubernetes RBAC governs cluster and pod access

Even when the database sits behind an internal load balancer, engineers still manage the pods through the Kubernetes API. RBAC refers to resources by the same name that appears in the API URL. Some APIs use a subresource, such as pod logs or pod command execution, written with a slash.

If a Role allows the pods exec subresource, a user can open a shell inside the Weaviate container. From there, they can read local files or inspect memory on the node. If a Role allows reading Secret objects, that user can pull the raw API keys used by the vector database.

Teams often grant wide ClusterRoleBindings during setup to fix deployment errors fast. Because native rules have no deny option, an overly broad binding gives permanent control over the namespace. You must scope each RoleBinding to the single namespace where the vector database runs.

Helm Chart Security + AKS Managed Identity: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

What native network and cluster controls leave open

An internal load balancer and namespace roles solve the first layer of security. The private IP keeps internet traffic out of your Azure subnet. Kubernetes RBAC decides which engineers can update the Helm release or view pod status.

Yet neither control inspects what happens inside an active database session or pod command. Network rules check where a packet comes from, not who sent the query or what data it returns. Once an app or user connects to the internal IP, the load balancer passes every request through.

Similarly, Kubernetes RBAC checks if a user can start a port-forward or exec session, not which commands run next. If vectors hold sensitive text like emails, phone numbers, card numbers, or national IDs, queries return them in plain text. Masking that rewrites matched fields in results is missing from native cloud load balancers and cluster roles.

Govern kubectl and API calls in flight

Identity providers and Kubernetes RBAC decide who logs into a cluster. They cannot redact keys inside returned JSON objects or hold a risky delete call for review. By design, hoop.dev runs as an open-source proxy, a sidecar for your Kubernetes API, checking requests and responses without replacing RBAC.

First, hoop.dev guardrails refuse banned methods, paths, or headers before they run. Next, hoop.dev data masking rewrites sensitive keys in returned Secret, ConfigMap, and log responses in memory, and writes each decision to the audit trail. Explore the repo on GitHub.

Frequently asked questions

Why use an internal load balancer for Weaviate on AKS?

An internal load balancer assigns a private IP from your Azure virtual network to the Weaviate service. This lets apps in peered networks or on-premises links query vectors without opening a public IP to the internet.

How do Kubernetes Roles limit access to a Weaviate Helm deployment?

A Kubernetes Role defines allowed actions inside a single namespace, and a RoleBinding assigns them to specific users. Because permissions are additive with no deny rules, teams must avoid broad ClusterRoleBindings and grant only needed verbs.

Can an internal load balancer mask sensitive fields in vector queries?

No, an Azure internal load balancer only routes network packets across private subnets. It does not inspect database queries, verify user identity on each request, or rewrite fields like emails or card numbers in results.

Short role checks help teams work fast. Simple rules make each review clear. Good logs keep every team safe. Scoped keys stop bad changes early.

Teams test new role rules in staging first. Clear owner lists speed up each audit. Small role scopes cut down on risk. Regular log checks catch drift right away.

The query runs. The emails stay hidden.