Incident brief
Invalid catalog name
A connection attempt named a database that doesn't exist. PostgreSQL rejected the connection before it was ever established.
In 10 seconds
- What triggers it
- Try to connect to a database name that has never been created.
- Fix
- Connect using the correct, existing database name.
- Proof
- Reproduced on PostgreSQL 16.14 → Connecting to a database name that had never been created was rejected with SQLSTATE 3D000. The correct database name, tried right afterward, connected normally.
Fix
What to do right now
Application-level steps for this error.
- Connect using the correct, existing database name.
- If the database genuinely should exist, create it first with CREATE DATABASE, but see the counterexample for why you can't do that inline as part of the same failed attempt.
- Double-check for typos and case, PostgreSQL folds unquoted database names to lowercase, so case mismatches trigger this same SQLSTATE even for a database that does exist.
-- list databases you can connect to, then reconnect with a real name
SELECT datname FROM pg_database WHERE datallowconn ORDER BY 1;
-- create only if it should exist (from a session already connected elsewhere):
-- CREATE DATABASE app_db;
-- reconnect: psql -d app_db
-- unquoted names fold to lowercase, MyDB and mydb are the same targetFor this error
See this error live on the server
Run these against the affected instance to confirm the diagnosis before you act.
3D000 occurs before application SQL runs. Confirm the requested database name, connect privilege, and whether a pooler is rewriting the target.
Databases this server can accept
Compare the exact database from the connection string with catalog spelling and allow_connections.
SELECT datname, datallowconn,
has_database_privilege(current_user, datname, 'CONNECT') AS can_connect
FROM pg_database
ORDER BY datname;Why it happens
What PostgreSQL is telling you
The mechanism behind the error, grounded in the official manual, not paraphrased.
PostgreSQL 16 Documentation, Creating a Database
In order to create a database, the PostgreSQL server must be up and running... Databases are created with the SQL command CREATE DATABASE.Read the full section on postgresql.org →
Nonexistent database name
analytics_replica has never been created in this cluster. PostgreSQL rejects the connection outright at startup, it never gets far enough to run a query.Correct database name, right afterward
The identical connection attempt, with only the database name corrected, succeeded immediately, confirming the earlier failure was specifically about the database name.Reproduce & verify
A real reproduction at the connection stage
A literal transcript of SQL run against a live PostgreSQL instance in an isolated lab. The commands below are exactly what was executed.
- 1Try to connect to a database name that has never been created.
- 2SQLSTATE 3D000 is reported as a FATAL connection error; no session is ever established.
One client, tried twice: first against a nonexistent database name, then against the correct one.
-- No schema needed: this reproduces at the connection stage, before any query runs.
SELECT 'no schema required' AS setup_note;psql -U postgres -d analytics_replica-- reconnect using the correct, existing database name
SELECT 1 AS connected_successfully;What PostgreSQL actually returned
setup_note
--------------------
no schema required
(1 row)psql: error: connection to server on socket "..." failed: FATAL: database "analytics_replica" does not exist connected_successfully
-------------------------
1
(1 row)A database created with a quoted, mixed-case name behaves differently from an ordinary lowercase one. This was tested directly: creating "BillingDB" (quoted, mixed case), then connecting with the plain lowercase form of that same name.
Without this
Above: guessing at a database name that was never created at all fails with SQLSTATE 3D000.
With this, tested
Below: a database that does exist can still trigger the identical SQLSTATE 3D000 if the case doesn't match.
- A second operational test: exact SQL, raw output, measured result, and engineer notes
- Fix that fails with a different, more confusing error: exact SQL, output, and verdict
- A manual-grounded production interpretation of the lab result
Card required. Cancel before day 7 and you are not charged.
Runbook to fix this
The runbook for this incident
Full step-by-step fixes for the condition behind this error: the diagnosis, the exact SQL, and output captured in the lab.
Connected
Everything this error touches
Every page this SQLSTATE connects to: the concept that explains it, the runbooks that fix it, the parameters you tune to prevent it, and the sibling errors it travels with. All real cross-references. Jump straight in, or open the full interactive map.
Fix it — runbooks
Related errors
Verification
- Last verified
- 2026-07-16 (Docker lab, PostgreSQL 16.14)
- Verification scope
- Verified against PostgreSQL 16.14 in an isolated lab environment
- Audit status
- reviewed
Went further?
Pro unlocks the second lab proof
Free page stops the bleeding. Pro adds the operational test, SQLSTATE audit, and deeper evidence, same error, more certainty.