Skip to content

Conversation

@levkk
Copy link
Collaborator

@levkk levkk commented Jan 16, 2026

Starting point to find a race condition in autocommit mode.

@levkk
Copy link
Collaborator Author

levkk commented Jan 16, 2026

@mkreidenweis-schulmngr I wrote this quick to help us reproduce the error you've been seeing. Let me know if you find something on your side.

@codecov
Copy link

codecov bot commented Jan 16, 2026

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

@mkreidenweis-schulmngr
Copy link

Cool. Don't have much time right now. But I guess an important factor is that we are running other queries on the database in parallel that are actually using explicit transactions with varying isolation levels (modifying unrelated tables).

@mkreidenweis-schulmngr
Copy link

Just wanted to let you know that I played around with this a bit, e.g. adding parallel transactions, but I also couldn't reproduce the effect so far with this demonstrator either.

When deploying PgBouncer for now, instead of PgDog, between our App Servers and database, we did not experience the issues described in #685. We we're even more convinced it's something within PgDog, not within our App Server, even if we couldn't reproduce it on a dev machine so far.

We'll probably look into this again soon.

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