Skippr Docs

Appearance

How to Configure MySQL as a Source Connector ​

May 2026

MySQL source setup is easy to keep clean if the connection string and table scope are decided before you scale the extract.

Short Answer ​

Configure MySQL as a source by setting source.kind to mysql, then supplying connection_string in the standard form mysql://user:pass@host:3306/db. If you only want a subset of the database, add tables; otherwise Skippr can auto-discover readable tables and ingest from the broader schema.

A clean first setup usually means exporting MYSQL_CONNECTION_STRING, pointing it at one known database, and choosing whether to start with explicit tables like customers,orders or let discovery show what the user can read. That gives you a simple way to validate both connectivity and table visibility before deciding whether the source should later run in CDC mode from the binlog.

Why Teams Struggle with This ​

Most MySQL configuration issues come from boundary confusion rather than from the connector itself. The URI may point at the right host but the wrong database, or the login may connect successfully while still lacking permission to inspect the tables your pipeline expects.

  • The connection string should be supplied through MYSQL_CONNECTION_STRING rather than stored inline.
  • The MySQL user needs read access to the selected tables.
  • The runner must reach the MySQL host and port from its own network environment.
  • Choosing an explicit table list can make first-run validation easier than full discovery on a large database.

How Skippr Handles It ​

Skippr keeps MySQL source configuration intentionally small: one URI plus optional table scoping. That works well because the first thing teams usually need is a dependable read from operational tables, not a large amount of connector ceremony before they can validate one database.

The first successful run gives you a clear checkpoint. Skippr can authenticate, inspect the selected schema, and either auto-discover readable tables or ingest only the named ones, which makes it easy to compare the extracted shape against the source system.

  • Simple MySQL URI-based source configuration.
  • Optional explicit table lists when you want a narrower first extract.
  • Compatibility with broader auto-discovery when you want to scale the source later.
  • A direct starting point for future binlog CDC setups documented elsewhere.

What the First Useful Version Looks Like ​

Start with a small set of high-signal tables so you can check row counts and column typing before widening the connector to the rest of the database.

If the source will eventually move into CDC, keep the first batch setup close to the production database and credentials so later cutover work stays small.