Easily Boost Ubuntu Network Performance by Enabling TCP BBR

This tutorial shows you how to enable TCP BBR on Ubuntu. TCP BBR is a TCP congestion control algorithm developed by Google. It tackles the shortcomings of traditional TCP congestion control algorithms (Reno or CUBIC). According to Google, it can achieve orders of magnitude higher bandwidth and lower latency. TCP BBR is already being used on Google.com and YouTube servers, and the Internet Engineering Task Force (IETF) has been standardizing this algorithm Since July, 2017. BBR stands for Bottleneck Bandwidth and RTT.

BBR requires only changes on the sender side. You can enable TCP BBR on your Linux desktop to improve the overall web surfing experience. If you have a Linux web server, TCP BBR can achieve faster web page downloads for your website visitors.

Step 1: Check TCP Congestion Control Algorithms on Linux

By default, Linux uses the Reno and CUBIC congestion control algorithm. To check available congestion control algorithms, run the following command.

sysctl net.ipv4.tcp_available_congestion_control


net.ipv4.tcp_available_congestion_control = cubic reno

To check the current congestion control algorithm in use, run

sysctl net.ipv4.tcp_congestion_control


net.ipv4.tcp_congestion_control = cubic

Step 2: Make Sure You Have Linux Kernel 4.9 or Above

TCP BBR is supported by Linux since kernel version 4.9. Use the following command to check your Linux kernel version.

uname -r

Ubuntu 18.04 has kernel 4.13.


Ubuntu 20.04 uses kernel 5.4.


Ubuntu 22.04 uses kernel 5.15.


Step 3: Enable TCP BBR in Ubuntu

Once you have kernel 4.9 or above, edit sysctl.conf file.

sudo nano /etc/sysctl.conf

Add the following two lines at the end of the file.


ubuntu tcp bbr

Save and close the file. Then reload sysctl configurations.

sudo sysctl -p

If you have correctly added the above two lines in the file, then they will be included in the output like below.

linuxbabe@ubuntu:~$ sudo sysctl -p
net.core.default_qdisc = fq
net.ipv4.tcp_congestion_control = bbr

Now check the congestion control algorithm in use.

sysctl net.ipv4.tcp_congestion_control


net.ipv4.tcp_congestion_control = bbr

Congrats! You have successfully enabled TCP BBR on Ubuntu.

Speed Improvements

I used Shadowsocks proxy on a low-end VPS to test TCP BBR. The speed improvement is huge. The following screenshot shows YouTube connection speed before enabling TCP BBR on my proxy server. As you can see, the speed is only around 400Kb/s (50KB/s).

enable TCP BBR on Ubuntu 16.04

This is YouTube connection speed after enabling TCP BBR on my proxy server. 9878 Kb/s is about 1234 KB/s. My proxy server just got 24 times faster!

enable google tcp bbr ubuntu

On a second Shadowsocks proxy server, the speed went from around 10Mbit/s to around 62Mbit/s (64172Kbit/s) after enabling TCP BBR.

TCP BBR performance boost

Don’t disable TCP Window Scaling

You might need to disable TCP window scaling in order to fix a problem, but doing so can greatly reduce the network performance. Add the following line in /etc/sysctl.conf file to enable TCP window scaling.

net.ipv4.tcp_window_scaling = 1

Then apply the changes.

sudo sysctl -p

That’s it! I hope this tutorial helped you enable TCP BBR on Ubuntu. As always, if you found this post useful, then subscribe to our free newsletter to get more useful tips and tricks.

Performance Monitoring

Rate this tutorial
[Total: 37 Average: 4.4]

15 Responses to “Easily Boost Ubuntu Network Performance by Enabling TCP BBR

  • MrAdminus
    2 weeks ago

    You should write a little notice on Linuxtoday so more people get to know about your blog. I think what you’re doing is great and contributes to the Linux community. It’s free to write a post there too and many Linux users reads it!

  • Greya Being
    2 weeks ago

    How do I undo this?

  • Sayyed Alireza
    3 months ago

    Best Blog website ever

    thanks for being born :*

  • Michael
    5 years ago

    Thank you for this little gem! So easy to do too 🙂

  • Hi
    I have tried to enable it but my output is different from yours. Below is the output from my system

    jieru@MS-7B51:~$ sysctl net.ipv4.tcp_available_congestion_control
    net.ipv4.tcp_available_congestion_control = reno cubic
    jieru@MS-7B51:~$ sysctl net.ipv4.tcp_congestion_control
    net.ipv4.tcp_congestion_control = cubic
    jieru@MS-7B51:~$ sudo nano /etc/sysctl.conf
    [sudo] password for jieru:
    jieru@MS-7B51:~$ sudo sysctl -p
    jieru@MS-7B51:~$ sysctl net.ipv4.tcp_congestion_control
    net.ipv4.tcp_congestion_control = cubic

    • Xiao Guoan (Admin)
      3 years ago


      If you have correctly added the above two lines in the file, then they will be included in the output like below when you run sudo sysctl -p.

      linuxbabe@ubuntu:~$ sudo sysctl -p
      net.core.default_qdisc = fq
      net.ipv4.tcp_congestion_control = bbr
      • Hi
        Here is the updated outputs

        jieru@MS-7B51:~$ sudo cat /boot/config-$(uname -r) | grep 'CONFIG_TCP_CONG_BBR'
        [sudo] password for jieru:           
        jieru@MS-7B51:~$ sudo cat /boot/config-$(uname -r) | grep 'CONFIG_NET_SCH_FQ'
        jieru@MS-7B51:~$ sudo nano /etc/sysctl.conf
        jieru@MS-7B51:~$ sudo sysctl --system
        * Applying /etc/sysctl.d/10-console-messages.conf ...
        kernel.printk = 4 4 1 7
        * Applying /etc/sysctl.d/10-ipv6-privacy.conf ...
        net.ipv6.conf.all.use_tempaddr = 2
        net.ipv6.conf.default.use_tempaddr = 2
        * Applying /etc/sysctl.d/10-kernel-hardening.conf ...
        kernel.kptr_restrict = 1
        * Applying /etc/sysctl.d/10-link-restrictions.conf ...
        fs.protected_hardlinks = 1
        fs.protected_symlinks = 1
        * Applying /etc/sysctl.d/10-magic-sysrq.conf ...
        kernel.sysrq = 176
        * Applying /etc/sysctl.d/10-network-security.conf ...
        net.ipv4.conf.default.rp_filter = 2
        net.ipv4.conf.all.rp_filter = 2
        * Applying /etc/sysctl.d/10-ptrace.conf ...
        kernel.yama.ptrace_scope = 1
        * Applying /etc/sysctl.d/10-zeropage.conf ...
        vm.mmap_min_addr = 65536
        * Applying /usr/lib/sysctl.d/50-coredump.conf ...
        kernel.core_pattern = |/lib/systemd/systemd-coredump %P %u %g %s %t 9223372036854775808 %h
        * Applying /usr/lib/sysctl.d/50-default.conf ...
        net.ipv4.conf.default.promote_secondaries = 1
        sysctl: setting key "net.ipv4.conf.all.promote_secondaries": Invalid argument
        net.ipv4.ping_group_range = 0 2147483647
        net.core.default_qdisc = fq_codel
        fs.protected_regular = 1
        fs.protected_fifos = 1
        * Applying /usr/lib/sysctl.d/50-mint.conf ...
        fs.inotify.max_user_watches = 65536
        * Applying /usr/lib/sysctl.d/50-pid-max.conf ...
        kernel.pid_max = 4194304
        * Applying /etc/sysctl.d/99-sysctl.conf ...
        net.core.default_qdisc = fq
        net.ipv4.tcp_congestion_control = bbr
        * Applying /usr/lib/sysctl.d/protect-links.conf ...
        fs.protected_fifos = 1
        fs.protected_hardlinks = 1
        fs.protected_regular = 2
        fs.protected_symlinks = 1
        * Applying /etc/sysctl.conf ...
        net.core.default_qdisc = fq
        net.ipv4.tcp_congestion_control = bbr
        jieru@MS-7B51:~$ sysctl net.ipv4.tcp_congestion_control
        net.ipv4.tcp_congestion_control = bbr

        Is everything correct?


    • Xiao Guoan (Admin)
      3 years ago

      Yes. it’s correct.

  • i did not understand what net.ipv4.tcp_window_scaling = 1 means

    • Xiao Guoan (Admin)
      3 years ago

      If you don’t understand this, then ignore it. It’s on by default. Learn the TCP/IP protocol and you will understand TCP window scaling.

  • Bruno Cabral
    2 years ago

    Is it necessary to install hwe kernel? Because Ubuntu 18.04 already has kernel > 4.9 (4.15, as today). HWE kernel would be 5.4

    • Xiao Guoan (Admin)
      2 years ago

      If you have Linux kernel 4.9 or above, then you don’t need to install a new kernel.

  • Hi Xiao, there is any booster for UDP like TCP as well??

  • Hi
    Tried installing softether on linux once more (currently running on a windows very old machine yet it runs) and i m having the same speed problem again ….. connected one client and results

    client speed results:
    54 Mbit Down / 7.2 Mbit Up 61ms softether on windows machine
    1.6 Mbit Down / 7.7 Mbit Up 76ms softether on linux machine

    Since linux is on a VM, added
    net.core.default_qdisc = fq
    net.ipv4.tcp_congestion_control = bbr

    What else to do? Is it because ens18 (internal naming for ethernet) is based on a vmbr which is based on 2 port bonding in LACP mode?

    During configuration I had a message about Promiscuous mode (I ll upload it). Can I do something with that?

Leave a Comment

  • Comments with links are moderated by admin before published.
  • Your email address will not be published.
  • Use <pre> ... </pre> HTML tag to quote the output from your terminal/console.
  • Please use the community (https://community.linuxbabe.com) for questions unrelated to this article.
  • I don't have time to answer every question. Making a donation would incentivize me to spend more time answering questions.

The maximum upload file size: 2 MB. You can upload: image. Links to YouTube, Facebook, Twitter and other services inserted in the comment text will be automatically embedded. Drop file here