Twingate Helm Chart

Published
Length
3 min read

Luis Felipe Dias

All posts

You deploy the twingate helm chart to run private network connectors inside your Kubernetes cluster. You supply your account tokens, and the chart starts connector pods that route user traffic. Teams often use this setup to replace old virtual private networks for cluster admins. Still, once a user connects to the network, broad cluster roles often grant too much power. How the twingate helm chart works in a cluster A connector is a small piece of software that runs inside your private networ

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.

You deploy the twingate helm chart to run private network connectors inside your Kubernetes cluster. You supply your account tokens, and the chart starts connector pods that route user traffic. Teams often use this setup to replace old virtual private networks for cluster admins. Still, once a user connects to the network, broad cluster roles often grant too much power.

How the twingate helm chart works in a cluster

A connector is a small piece of software that runs inside your private network. It opens an outbound tunnel to the central control plane over standard web ports. Because it only makes outbound calls, you do not need to open inbound firewall ports. The chart packages this connector software so Kubernetes can run and manage it for you.

Before you run the chart, you create a connector record in your admin console. The console gives you two secret tokens that identify and authorize that specific connector instance. You pass those tokens into the chart values along with your unique network name. It is best to store those tokens in a Kubernetes secret before deployment begins.

When you install the chart, Kubernetes schedules the connector pod on a worker node. The pod reads the secret tokens and registers itself with the remote relay servers. Teams usually run at least two connector pods in separate zones for high uptime. If one node fails, the second pod keeps the network tunnel open for your engineers.

Managing updates and cluster resources

Running connectors through a chart makes day-to-day maintenance much easier for platform teams. You can pin the connector image version in your values file to control rollouts. When a new version comes out, you update the chart and run a rolling upgrade. One pod updates at a time so active user sessions do not drop during changes.

You can also set memory and CPU limits in the chart to protect your nodes. The connector pods can use node affinity rules to stay on dedicated system machines. Some teams also deploy a separate operator chart to sync network resources through Kubernetes manifests. Either way, the connector pods act only as a transport bridge into your private subnet.

Helm Chart Security: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

What network connectors solve and what they miss

This setup solves the problem of exposing private cluster endpoints to the public internet. Only users who log in through your identity provider can reach the internal network IP. You can restrict which groups can reach the Kubernetes API or internal database ports. This stops outside attackers from scanning your ports or guessing passwords from the internet.

Still, a network connector only decides who can open a socket to a port. Once the tunnel connects, the network tool steps out of the way of the traffic. It does not read the commands sent to the cluster or the queries run on databases. Users still need local database passwords or static cluster keys on their own laptops.

This leaves a big gap during daily work and security audits. If an engineer has a shared database role, they can run a drop table statement. They can also query tables full of customer emails, phone numbers, and card numbers. The network log only shows that a socket opened, not what commands ran inside it.

Check Kubernetes API requests and responses on the wire

Kubernetes RBAC decides who may call the API server. RBAC does not check response payloads or split listing Secret names from reading Secret values when a role grants get access. With hoop.dev, you run an open-source HTTP proxy, a sidecar for your Kubernetes API. The kubectl client changes only the server address it talks to while RBAC stays in place.

On each call, hoop.dev guardrails read the method, normalized path, and headers to refuse bad deletes or full Secret reads with a 403 error. On the way back, hoop.dev data masking walks JSON responses by key and masks matched values in Secrets, ConfigMaps, and pod logs. Every request and verdict is logged without storing bearer tokens. View the project on GitHub.

Frequently asked questions

How many connector pods should I run per cluster?

You should run at least two connector pods in each cluster for high availability. Place them on different nodes so node maintenance does not cut off your team.

Do I need to open inbound firewall ports for the chart?

No, the connector pods only make outbound connections to the cloud relay servers. You can keep all inbound firewall ports closed on your cluster network.

Does the connector log the SQL queries or kubectl commands my team runs?

No, the connector operates at the network transport layer and only routes packets to ports. It does not inspect, log, or block specific database queries or cluster commands.

The query runs. The emails stay hidden.