Nikita Kazmin

Hello. Yeah, I agree with you that it looks like a keepalive issue. I have one particular place in mind, I'll take a look. It would also be great if...

OK, for now I can confirm that there is indeed an issue while calling `NpgsqlConnection.ReloadTypes` with keepalive (that's the first and the third stack traces). We do not call `StartUserAction`...

Sorry, for the wait, was busy with other stuff. I think it's OK to merge this.

Note that users can already do all of this via [processors](https://github.com/open-telemetry/opentelemetry-dotnet/blob/main/docs/trace/extending-the-sdk/README.md#processor).

~~Now that's curious. @roji any ideas?~~ The latest postgres release changed things...

Backported to 6.0.7 via bca2210.

Hello, sorry for the wait, this issue slipped under the radar. The problem here is that parameter's names differ from the ones procedure has (that's the ones PG sends to...

Hello, there was a similar issue raised no so long ago (#3942). Most of the time it happens whenever there is a problem on pg's side - the best thing...

@BrunoSeixas if you're using the official postgres image you might want to raise this issue with [them](https://github.com/docker-library/postgres), since it doesn't look like npgsql is at fault here.

I've done a quick test, and these are the results I've got. Npgsql 2.2.7: ``` ini BenchmarkDotNet=v0.13.1, OS=Windows 10.0.19044.1826 (21H2) AMD Ryzen 5 5600X, 1 CPU, 12 logical and 6...