Airbyte Helm Chart
You deploy the airbyte helm chart by installing the chart package on a Kubernetes cluster. This setup runs your data sync pods, web app, and job schedulers in one namespace. Most platform teams start with a basic setup to move data between sources and destinations. Still, teams often grant broad cluster roles and static database keys just to make sync jobs run. Those wide permissions create major risks when engineers need to check or fix broken sync pods. How the airbyte helm chart architectur
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 airbyte helm chart by installing the chart package on a Kubernetes cluster. This setup runs your data sync pods, web app, and job schedulers in one namespace. Most platform teams start with a basic setup to move data between sources and destinations. Still, teams often grant broad cluster roles and static database keys just to make sync jobs run. Those wide permissions create major risks when engineers need to check or fix broken sync pods.
How the airbyte helm chart architecture works
The airbyte helm chart bundles several core services into a single Kubernetes deployment. A web server runs the user interface and handles API calls from your data team. A scheduler pod tracks sync timing and starts worker pods for each data transfer job. Each worker pod spins up temporary containers that read from sources and write to targets. A backing database stores job state, sync history, and source keys for the whole app. Object storage holds job logs so engineers can read output when a sync fails. Kubernetes service accounts give the scheduler permission to create and delete job pods in the namespace.
Managing credentials and cluster roles in Kubernetes
Teams must set up secrets for the internal database and external cloud storage buckets. The chart reads these Kubernetes secrets and mounts them into the server and worker pods. You also set up ingress rules so users can log in to the web app. Many teams link the web interface to a single sign-on provider for user logins. At the cluster level, engineers use role-based rules to control who can manage the namespace. A standard role lets an engineer view pod status or restart a stuck sync worker. Still, engineers often need direct access to the internal database or cluster shell to fix issues.
What native setup solves and what it leaves open
Native Kubernetes roles and chart secrets handle basic app startup and pod scheduling well. They decide which service account can launch worker pods and which users can log in. Still, native chart settings only check identity at the start of a session. Once an engineer opens a shell or database session, the cluster does not check single commands. A broad role lets a user run risky queries or read sensitive data in flight. Static secrets stored in the namespace also stay exposed to anyone with pod read permissions. If a sync job pulls sensitive customer emails or card numbers, those fields sit in logs. Native cluster controls do not mask sensitive fields like phone numbers or national IDs in query results. Teams need a way to inspect commands and protect data during live debugging sessions.
Helm Chart Security: architecture patterns and best practices
Continue reading? Get the full guide.
Free. No spam. Unsubscribe anytime.
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 sits inline as an open-source proxy, a sidecar for your Kubernetes API, without replacing RBAC. On every production call from an engineer or AI agent, hoop.dev decides to pass, mask, block, or hold it for human approval.
First, hoop.dev guardrails refuse banned methods, paths, or headers before they run or hold risky deletes for review. 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
How should we handle the internal database in production?
Do not use the default in-cluster database pod for real production data loads. Set up a managed external PostgreSQL instance and pass its secret to the chart instead.
Why do sync worker pods fail to start on the cluster?
Worker pods usually fail when the scheduler service account lacks permission to create pods. Check your namespace role bindings and make sure the node pool has enough memory.
How can we keep sensitive data out of sync job logs?
Limit who can read pod logs and external log buckets through strict role rules. Apply data masking to rewrite emails, phone numbers, and card numbers before engineers read results.
Related guides
In this post