How to enable TCP BBR support?

[Jones]

Code: [Select]

$ tce-load -w -i net-sched-5.4.3-tinycore.tcz
$ sudo su
# echo 'cake' > /proc/sys/net/core/default_qdisc
# echo 'bbr' > /proc/sys/net/ipv4/tcp_congestion_control
sh: write error: No such file or directory

No TCP BBR support?

[Juanito]

It exists on CorePure64 at least:

Code: [Select]

$ ls /proc/sys/net/ipv4/*congestion*
/proc/sys/net/ipv4/tcp_allowed_congestion_control
/proc/sys/net/ipv4/tcp_available_congestion_control
/proc/sys/net/ipv4/tcp_congestion_control
$ cat /proc/sys/net/ipv4/tcp_available_congestion_control
reno cubic
$ cat /proc/sys/net/ipv4/tcp_allowed_congestion_control
reno cubic
$ cat /proc/sys/net/ipv4/tcp_congestion_control
cubic



[Jones]

It exists on CorePure64 at least:

Code: [Select]

$ ls /proc/sys/net/ipv4/*congestion*
/proc/sys/net/ipv4/tcp_allowed_congestion_control
/proc/sys/net/ipv4/tcp_available_congestion_control
/proc/sys/net/ipv4/tcp_congestion_control
$ cat /proc/sys/net/ipv4/tcp_available_congestion_control
reno cubic
$ cat /proc/sys/net/ipv4/tcp_allowed_congestion_control
reno cubic
$ cat /proc/sys/net/ipv4/tcp_congestion_control
cubic



Only reno and cubic, no bbr algorithm by google, 

[Rich]

Hi Jones
Welcome to the forum.

To enable  BBR  requires  TCP_CONG_ADVANCED  to be set. The kernel config file shows:

Code: [Select]

# CONFIG_TCP_CONG_ADVANCED is not set
You would need to recompile the kernel if you want BBR.

[Rich]

Hi Jones
The help text for configuring  TCP_CONG_BBR  states:

It requires the fq ("Fair Queue") pacing packet scheduler.

I don't see anything in the Kconfig file that enables that when  TCP_CONG_BBR  is selected. You should check that:

Code: [Select]

NET_SCH_FQis enabled prior to compiling.

[Jones]

Hi Jones
The help text for configuring  TCP_CONG_BBR  states:

It requires the fq ("Fair Queue") pacing packet scheduler.

I don't see anything in the Kconfig file that enables that when  TCP_CONG_BBR  is selected. You should check that:

Code: [Select]

NET_SCH_FQis enabled prior to compiling.

Thanks, I will try it later.