Database guides

Presto

Updated 2026-08-01 · 2 min read

The engine Trino was forked from, served by the same AddisDB driver with Presto’s own header namespace.

Presto and Trino share a statement protocol — Trino is a fork of Presto — so AddisDB drives both with one driver, differing only in the header namespace each expects. Everything you get for Trino, you get for Presto.

Who it is for

PrestoDB is the original Facebook-built federated SQL engine, now under the Linux Foundation. It queries data where it lives across many connectors, and is the engine behind several managed services.

The split happened in 2020, and the two have drifted since: function names, connector features and SQL extensions no longer match one-for-one. Code written for one is usually close enough to run on the other, and occasionally not.

You will be on Presto rather than Trino if you inherited an existing deployment, if a vendor’s platform standardized on it, or if you are working against Amazon Athena’s Presto-based engine.

Set up the server

  1. Local: docker run -p 8080:8080 prestodb/presto.
  2. Configure connectors under etc/catalog/ — one properties file per underlying system.
  3. Note the coordinator host and port, and whether authentication is enabled.
  4. Confirm the cluster is up by visiting its web UI on port 8080.

Connect from AddisDB

  1. New Connection → Presto under Big Data / Warehouse. Port prefills to 8080 and the username to presto.
  2. Enter the coordinator host. The username is sent as Presto’s user header.
  3. Set Database to catalog or catalog/schema.
  4. Add a password in the API key / token field only if your cluster requires authentication.
  5. Test, then Save.

Pick the Presto engine rather than Trino even though the protocol is shared — the two expect differently namespaced headers, and a mismatch is rejected by the coordinator.

Finding your way around a cluster

-- What is actually connected to this coordinator?
SHOW CATALOGS;
SHOW SCHEMAS FROM hive;
SHOW TABLES FROM hive.sales;

-- Three-part names, exactly as in Trino
SELECT * FROM tpch.sf1.nation LIMIT 10;

What AddisDB gives you

  • The Big Data Console with a catalog browser, query pane and preview row cap.
  • Cost estimation via EXPLAIN (TYPE IO).
  • Paged result following, with queries released on the server when finished.
  • Full safety, the Chart view, notebooks, ⌘K search, and AI querying.
  • The Live Monitor over system.runtime.queries.