What Network Access Skippr Needs for MySQL

June 2026

MySQL source access starts with a reachable host and port in the connection string, then expands if you also need binlog-based CDC.

Short Answer

Skippr needs a MySQL connection string that resolves to a reachable host and port, plus a user with read access to the selected tables. The MySQL source docs define the connector around a DSN like mysql://user:pass@host:3306/db, so the useful network boundary is that exact host, port, and database rather than a loose claim that the runner can reach MySQL somewhere.

That matters because the first successful batch read and the later CDC path are not the same thing. A batch integration can work once the runner can open the MySQL connection and read customers or orders. If you later enable CDC, the connector still depends on the same network path, but the database also has to satisfy binlog replication prerequisites documented in the CDC guide.

Why Teams Struggle with This

MySQL source setups often look simple because the DSN is short. The trap is that a short connection string still hides the real operational questions: can the runner reach host:3306, does the user have table-read access, and is CDC being assumed before its prerequisites are ready?

• The connection string must point at a host and port reachable from the actual runner environment.
• The MySQL user needs read access to the tables you want to ingest, not just to the database login.
• A batch extract succeeding once does not prove binlog-based CDC is ready.
• If tables is over-specified or wrong, the source can look partially healthy while still missing the data you expected.

How Skippr Handles It

Skippr keeps MySQL access easy to reason about by using one source connection string and an optional table list. That lets you validate the network path with a deliberately small first read before you add the extra surface area that comes with continuous change capture.

It also gives you a clear handoff to infrastructure or DBAs: one hostname, one port, one database, one user, one short list of tables. If any part of that is unclear, the source is not actually ready yet.

• Uses a standard MySQL connection string such as mysql://user:pass@host:3306/db.
• Supports a narrow tables list or auto-discovery for broader batch reads.
• Separates the basic read path from the later CDC binlog path.
• Keeps credentials in an environment variable instead of committing them to config.

What the First Useful Version Looks Like

The first useful version is one MySQL DSN, one database, and a small readable table set that proves the host, port, and auth path are correct.

Only move to CDC after that path is boring. If the runner still cannot consistently read the tables, adding replication concerns will not help.