-
Notifications
You must be signed in to change notification settings - Fork 139
Open
Description
Hi
I'm testing TQuic under high load (around 1.5-2 Gbps rx traffic) and sometimes I got an InternalError during call to Endpoint.process_connections. The output in the log file:
ERROR tquic::endpoint] SERVER generate packet err: InternalError
Subsequent calls to process_connections produce the same error and it could continue up to 10 seconds (which is equal to the configured idle timeout, is it related?).
Is there any way to understand why this happens?
Also, I went through the code and I see that this error is triggered when endpoint's connections are being processed. And while it happens with a single connection the other endpoint's connections are left unprocessed, because the error ends the loop. Is it a correct way to handle this?
Metadata
Metadata
Assignees
Labels
No labels