Skip to content

Conversation

@staaldraad
Copy link
Member

What kind of change does this PR introduce?

Feature

What is the current behavior?

db_handler understands the authentication_cleartext_password message but doesn't have a handler. This means servers using PAM for authentication do not work with db_handler as the password is never sent in response.

What is the new behavior?

Adds a handler for authentication_cleartext_password

Additional context

This is needed for other projects that will use PAM to extend the auth mechanisms offered in the hosted postgres at Supabase

db_handler understands the authentication_cleartext_password message but
doesn't have a handler. This means servers using PAM for authentication
do not work with db_handler as the password is never sent in response.

Adds a handler for authentication_cleartext_password
@staaldraad staaldraad requested a review from a team as a code owner July 22, 2025 07:50
@v0idpwn v0idpwn enabled auto-merge (squash) July 22, 2025 15:18
@v0idpwn v0idpwn merged commit deaa481 into main Jul 22, 2025
12 checks passed
@v0idpwn v0idpwn deleted the etienne/sec-456-support-password-auth-in-pooler branch July 22, 2025 15:25
@v0idpwn v0idpwn mentioned this pull request Jul 28, 2025
v0idpwn added a commit that referenced this pull request Jul 29, 2025
### Features
- **Authentication cleartext password support** - Added support for
cleartext password authentication method (#707)
- **Runtime-configurable connection retries** - Support for runtime
configuration of connection retries and infinite retries (#705)
- **Enhanced health checks** - Check database and eRPC capabilities
during health check operations (#691)
- **More consistency with postgres on auth errors** - Improves errors in
some client libraries (#711)

### Performance Improvements

- **Optimized ranch usage** - Supavisor now uses a constant number of
ranch instances for improved performance and resource management when
hosting a large number of pools (#706)

### Monitoring

- **New OS memory metrics** - gives a more accurate picture of memory
usage (#704)
- **Add a promex plugin for cluster metrics** - for tracking latency and
connection status (#690)
- **Client connection lifetime metrics** - adds a metric about how long
each connection is connected for (#688)
- **Process monitoring** - Log when large process heaps and long message
queues (#689)

### Bug Fixes

- **Client handler query cancellation** - Fixed handling of
`:cancel_query` when state is `:idle` (#692)

### Migration Notes

- Instances running a small number of pools may see an increase in
memory usage. This can be mitigated by changing the ranch shard or the
acceptor counts.
- If using any of the new used ports, may need to change the defaults
- Review monitoring dashboards and include new metrics
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants