Daniel (dB.) Doubrovkine

[Catch All Triage - Attendees [1](https://github.com/dblock/), [2](https://github.com/krisfreedain), [3](https://github.com/andrross), [4](https://github.com/gkamat), [5](https://github.com/getsaurabh02)]

[Catch All Triage - Attendees [1](https://github.com/dblock/), [2](https://github.com/krisfreedain), [3](https://github.com/andrross), [4](https://github.com/gkamat), [5](https://github.com/getsaurabh02)]

@ArnaudParant what's the root cause? Is the server timing out causing the parallel requests to back up? It would be great to turn the repro into a failing test.

Don't lose the repro. It's likely a pool being held too long, log, buffering, or something that should have long been garbage collected. Try to add and get a flame...

Check what `parallel_bulk` is doing and try to modify it to narrow down what is being held onto? There should be a way to `yield` results incrementally and avoid holding...

Thanks for your help @ArnaudParant!

@dependabot recreate

Legit failures here that need looking. ``` 1) Slack::Web::Api::Endpoints::AdminUsersSession admin.users.session_invalidate requires user_id Failure/Error: expect { client.admin_users_session_invalidate(session_id: %q[12345]) }.to raise_error ArgumentError, /Required arguments :user_id missing/ expected ArgumentError with message matching /Required...

@gregsaab I usually do this one before every release, does it need to be? For file upload v2 I think it's really a utility method that wraps calling an API...

@reta pretty great stuff - there's other work on improving client/server performance, how do you see us fit work on Protobuf (@VachaShah) / gRPC?