[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Future of the TCP_CONGCTL socket option ?
On some systems it is possible to call getsockopt() with parameter
TCP_CONGESTION to get the congestion control algorithm currently in use
by a TCP connection. On NetBSD the constant seems to be called
TCP_CONGCTL, but it isn't enabled yet:
Apparently the modular TCP congestion control API was added in NetBSD
about 16 years ago ( https://github.com/NetBSD/src/commit/f333039 ),
with following comment: "In the near future, it will be possible to
selected a congestion control algorithm on a per-socket basis."
Did I miss something or this per-socket TCP CCA selection functionality
has never been actually enabled ? In this case, I guess just reading the
current value of "net.inet.tcp.congctl.selected" (resp.
"net.inet6.tcp6.congctl.selected") is enough to know the CCA used
by all current connections, except in the cases where the value was
changed after the sockets have been created ?
Main Index |
Thread Index |