Skip to content

RtpQueue is full with newCc after 65536 seconds (max uint32_t in ntp domain) #55

@realuptime

Description

@realuptime

Hi,

ScreamTx rust plugin constantly logs "RtpQueue is full" after running for 65536 seconds (65536 * 65536 in ntp domain = 4294967296 = max uint32_t). For sure it is not ok to keep ntp domain time in a uint32_t variable type.
It happens when no receiver is connected.
After this point, the receiver will not be able to receive any RTP packet.
BTW, is there a way to reset screamtx without recreating the pipeline or re-inserting a new instance of screamtx?
I have attached 2 logs when this happened:
repro_queue_full1.zip
repro_queue_full2.zip

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions