Skip to content

rtpjitterbuffer: don't send multiple instant RTX for the same packet

Håvard Graff requested to merge hgr/gst-plugins-good:hgr/rtx-fix into master

Due to us not properly acknowleding the time when the last RTX was sent when scheduling a new one, it can easily happen that due to the packet you are requesting have a PTS that is slightly old (but not too old when adding the latency of the jitterbuffer), both its calculated second and third (etc.) timeout could already have passed. This would lead to a burst of RTX requests, which acts completely against its purpose, potentially spending a lot more bandwidth than needed.

This has been properly reproduced in the test: test_rtx_not_bursting_requests

The good news is that slightly re-thinking the logic concerning re-requesting RTX, made it a lot simpler to understand, and allows us to remove two members of the RtpTimer which no longer serves any purpose due to the refactoring. If desirable the whole "delay" concept can actually be removed completely from the timers, and simply just added to the timeout by the caller of the API. But that can be a change for a another time.

The only external change (other than the improved behavior around bursting RTX) is that the "delay" field now stricly represents the delay between the PTS of the RTX-requested packet and the time it is requested on, whereas before this calculation was more about the theoretical calculated delay. This is visible in three other RTX-tests where the delay had to be adjusted slightly. I am confident however that this change is correct.

Merge request reports