How to Configure DuckDB Proxy

Published
Length
3 min read

Luis Felipe Dias

All posts

You set up a duckdb proxy by setting network rules in your local engine. This routes outbound traffic through a relay server to reach remote files. Data teams often run this in-process analytical engine on local laptops or servers. To read remote object storage, users load cloud keys into their local session. When those keys have broad roles, one user can read or delete everything. How a native duckdb proxy works in practice DuckDB runs inside your host app or local command line tool. Unlike

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 set up a duckdb proxy by setting network rules in your local engine. This routes outbound traffic through a relay server to reach remote files. Data teams often run this in-process analytical engine on local laptops or servers. To read remote object storage, users load cloud keys into their local session. When those keys have broad roles, one user can read or delete everything.

How a native duckdb proxy works in practice

DuckDB runs inside your host app or local command line tool. Unlike traditional database servers, it does not listen for incoming network traffic. Instead, the engine reaches out over the network to read remote files. Teams store these parquet or CSV files in cloud buckets or web servers. When a strict network blocks direct web traffic, the engine needs a relay. You set the host address and port inside the local session settings. You also supply a username and password if the network relay requires them. Once set up, the engine sends all remote file requests through that relay. The relay forwards the request to the cloud storage bucket and returns data. Your local processor then runs the analytical query on the returned data blocks. This workflow lets analysts query huge remote datasets from their own machines. They do not need to copy full files onto their local hard drives.

What native network settings solve and what they miss

Native network routing solves a basic connectivity problem inside locked-down corporate networks. It lets an isolated server fetch remote files through one approved network path. Network admins can watch traffic volume and filter which domains the host reaches. Still, an outbound network relay only checks where traffic goes, not what runs. To read cloud buckets, the user still needs storage keys on their machine. Those keys often grant full read access to entire folders of raw data. Once the storage system accepts the key, the engine pulls any file inside. The network relay cannot see the SQL query running inside the local process. It cannot tell which user ran a select query on a sensitive table. It also cannot mask personal data like emails, phone numbers, or card numbers. If a table holds national IDs, the query returns those raw values intact. Native settings decide which machine reaches the bucket, not what queries return.

How to plan roles and storage permissions

Before you share remote data, split your cloud buckets by data sensitivity. Keep public analytics tables in one folder and private records in another folder. Give your analysts read-only roles that match their daily job tasks. Never share one master storage key across a whole engineering team. When everyone shares one key, your storage logs cannot show who read what. Rotate your storage keys often and remove old keys as people switch teams. Check your bucket logs each week to spot strange download spikes early. Good storage rules reduce risk, even when queries run on local machines.

Database Proxy (ProxySQL, PgBouncer) + End-to-End Encryption: architecture patterns and best practices

Continue reading? Get the full guide.

Free. No spam. Unsubscribe anytime.

Govern statements and responses across databases and APIs

Good config and scoped roles are needed setup. They do not check single queries in flight or mask sensitive fields in returned rows. Across Postgres, MySQL, SQL Server, MongoDB, ClickHouse, Kubernetes APIs, SSH, and internal HTTP endpoints, hoop.dev sits inline in the protocol beside each supported service as an open-source proxy, a sidecar for your database or API. For every query or command an AI agent or engineer runs against production, hoop.dev decides to pass, mask, block, or hold it for human approval.

First, hoop.dev guardrails block banned queries before they run or hold risky ones for review, and hoop.dev data masking rewrites matched sensitive values in responses in memory. Furthermore, hoop.dev holds SOC 2 Type II and is open source on GitHub.

Frequently asked questions

Does DuckDB accept incoming client connections like Postgres?

No, it is an embedded engine that runs inside your local app process. It makes outbound requests to read files rather than taking incoming database connections.

Why do analysts use a network relay with an embedded database?

Private servers often lack direct internet routes to public cloud storage buckets. A network relay gives the engine a safe path to fetch remote files.

Can outbound network settings mask sensitive columns in query results?

No, basic network routing only forwards raw file bytes from storage to memory. Masking fields like emails or card numbers requires inspecting the query results directly.

The query runs. The emails stay hidden.