[Bug]: Postgres: Connection failed

[markyarnold]

Describe the bug

Creating new Postgres connection using SSH Tunnel which Tests fine. Testing the RDS connection throws the error:

Connection failed: Error occurred while creating a new object: db error.

This was working find on the previous PG connector which is now deprecated. The new one, which I migrated to fails for all existing connections and new ones also.

To Reproduce

1. Create new PostgreSql connection
2. Select via SSH
3. Test SSH Connection - pass
4. Test RDS connection - Fail with the error - Connection failed: Error occurred while creating a new object: db error
5. Check username/password and role. Master username being used. Admin role. All grants.
6. Check same creds / setup with other DB tool - success.

OS Version

macOS 27

Tabularis Version

v0.w4.0

Relevant Log Output

Diagnostics:
Connection failed

09:01:18  Opening SSH tunnel (ec2-user@35.178.128.112:22)
09:01:18  SSH tunnel established (127.0.0.1:57549)
09:01:18  Connecting to the database (policy-management-platform-aurora-instance-prod.cd2yqwquoud3.eu-west-2.rds.amazonaws.com:5432)
09:01:18  Database connection failed (Connection failed: Error occurred while creating a new object: db error)

Connection failed: Error occurred while creating a new object: db error

[debba]

Thanks for the feedback!
I will transfer the issue in the right repo

[aesslinger]

@markyarnold Thanks for the report. This looks like the generic deadpool fallback error message (Error occurred while creating a new object: db error) rather than the real PostgreSQL error — a pattern we fixed in this plugin (issues #66/#67) so the actual server-side message (e.g. wrong password, database not found, etc.) surfaces instead. That fix has been in every release since v1.0.0-rc.2.

Could you confirm which Tabularis app version and PostgreSQL plugin build you're on, upgrade to the latest (v1.0.0-rc.4 or newer) if you haven't, and retry the RDS connection? If it still fails, please paste the fresh diagnostics/log block like you did above — with the fix in place it should show the actual Postgres-side reason instead of "db error", which will tell us exactly what's wrong.

[markyarnold]

@aesslinger Just checked and the plugin was at the earlier version - I was sure I had updated that. Have just done it, and the issue resolved. Thanks so much for your time; sorry it was wasted. All the best.

[aesslinger]

@markyarnold Thanks for the quick response and for verifying! Glad to hear updating to the latest plugin resolved it — and no time wasted at all, this kind of report helps confirm the fix is working as intended.