[ISSUE] linux: tcp congestion and optimizations

[ISSUE] linux: tcp congestion and optimizations

[tetsumaki]

[-- Attachment #1: Type: text/plain, Size: 1499 bytes --]

New issue by tetsumaki on void-packages repository

https://github.com/void-linux/void-packages/issues/39951

Description:
Hello everyone,

I've been using Void Linux for a long time now, and I found that the kernel was not optimized in an optimal way.

Void Linux uses the following settings:

```ini
net.ipv4.tcp_available_congestion_control = reno
net.core.default_qdisc = pfifo_fast
net.ipv4.tcp_congestion_control = reno
```

Most distributions (Rocky/Arch/Alpine) use the following settings:

```ini
net.ipv4.tcp_available_congestion_control = cubic bbr
net.core.default_qdisc = fq_codel
net.ipv4.tcp_congestion_control = cubic
```

I think the best setting though is this (same as above):

```ini
net.ipv4.tcp_available_congestion_control = cubic bbr
net.core.default_qdisc = fq_codel
net.ipv4.tcp_congestion_control = cubic
```

BBR is very good, but still in alpha, you have to give the user the choice to apply it.

Whatever happens, the kernel should be optimized, and the user should be allowed to change via sysctl, which is not the case now.

The discussion is open (fq/fq_codel/sch_fq/cak, reno/bbr and others optimizations).

My english is not perfect ;)

Some links :
- https://github.com/google/bbr/blob/master/README
- https://wiki.archlinux.org/title/sysctl#Enable_BBR
- https://www.cyberciti.biz/cloud-computing/increase-your-linux-server-internet-speed-with-tcp-bbr-congestion-control/
- https://github.com/systemd/systemd/issues/9725

Re: linux: tcp congestion and optimizations

[sgn]

[-- Attachment #1: Type: text/plain, Size: 513 bytes --]

New comment by sgn on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279167266

Comment:
> BBR is very good, but still in alpha, you have to give the user the choice to apply it.
> 
> Whatever happens, the kernel should be optimized,
> and the user should be allowed to change via sysctl,
> which is not the case now.

What is your problem when you're changing?
```
# modprobe tcp_bbr
# sysctl -w net.ipv4.tcp_congestion_control=bbr
```

works for me

Re: linux: tcp congestion and optimizations

[tetsumaki]

[-- Attachment #1: Type: text/plain, Size: 228 bytes --]

New comment by tetsumaki on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279193185

Comment:
Hi, I think the default configuration should be changed to use cubic and not reno.

Re: linux: tcp congestion and optimizations

[sgn]

[-- Attachment #1: Type: text/plain, Size: 507 bytes --]

New comment by sgn on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279195012

Comment:
If you want to do it automatically at boot:

```sh
# cat /etc/modules-load.d/tcp-congestion.conf
tcp_bbr
# cat /etc/sysctl.d/tcp-congestion.conf
net.ipv4.tcp_available_congestion_control=bbr
```



> Hi, I think the default configuration should be changed to use cubic and not reno.

That's subjective. You have the option to change to anything you want.

Re: linux: tcp congestion and optimizations

[tetsumaki]

[-- Attachment #1: Type: text/plain, Size: 226 bytes --]

New comment by tetsumaki on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279216352

Comment:
Yes, i already use fq_codel / bbr.
I'm talking about the default configuration.

Re: linux: tcp congestion and optimizations

[sgn]

[-- Attachment #1: Type: text/plain, Size: 372 bytes --]

New comment by sgn on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279221730

Comment:
> Yes, i already use fq_codel / bbr. I'm talking about the default configuration.

I don't understand, your first message is "Whatever happens, the kernel should be optimized, and the user should be allowed to change via sysctl."?

Re: linux: tcp congestion and optimizations

[tetsumaki]

[-- Attachment #1: Type: text/plain, Size: 490 bytes --]

New comment by tetsumaki on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1279275066

Comment:
> > Yes, i already use fq_codel / bbr. I'm talking about the default configuration.
> 
> I don't understand, your first message is "Whatever happens, the kernel should be optimized, and the user should be allowed to change via sysctl."?

This is a bad translation, my bad.
And the module was not loaded, so it was not visible. on my paste.

Re: linux: tcp congestion and optimizations

[sgn]

[-- Attachment #1: Type: text/plain, Size: 266 bytes --]

New comment by sgn on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1280119053

Comment:
I changed congestion control to cubic, it will available through new linux5.19 and linux6.0
Other kernels will follow suites.

Re: [ISSUE] [CLOSED] linux: tcp congestion and optimizations

[sgn]

[-- Attachment #1: Type: text/plain, Size: 1853 bytes --]

Closed issue by tetsumaki on void-packages repository

https://github.com/void-linux/void-packages/issues/39951

Description:
Hello everyone,

I've been using Void Linux for a long time now, and I found that the kernel was not optimized in an optimal way.

Void Linux uses the following settings:

```ini
net.ipv4.tcp_available_congestion_control = reno
net.core.default_qdisc = pfifo_fast
net.ipv4.tcp_congestion_control = reno
```

Most distributions (Rocky/Arch) use the following settings:

```ini
net.ipv4.tcp_available_congestion_control = cubic bbr
net.core.default_qdisc = fq_codel
net.ipv4.tcp_congestion_control = cubic
```

Alpine use the following settings:

```ini
net.ipv4.tcp_available_congestion_control = reno cubic
net.core.default_qdisc = pfifo_fast
net.ipv4.tcp_congestion_control = cubic
```

I think the best setting though is this (Rocky/Arch):

```ini
net.ipv4.tcp_available_congestion_control = cubic bbr
net.core.default_qdisc = fq_codel
net.ipv4.tcp_congestion_control = cubic
```

BBR is very good, but still in alpha, you have to give the user the choice to apply it.

Whatever happens, the kernel should be optimized, and the user should be allowed to change via sysctl.

The discussion is open (fq/fq_codel/sch_fq/cak, reno/bbr and others optimizations).

My english is not perfect ;)

Some links :
- https://github.com/google/bbr/blob/master/README
- https://wiki.archlinux.org/title/sysctl#Enable_BBR
- https://www.cyberciti.biz/cloud-computing/increase-your-linux-server-internet-speed-with-tcp-bbr-congestion-control/
- https://github.com/systemd/systemd/issues/9725
- https://raw.githubusercontent.com/archlinux/svntogit-packages/packages/linux/trunk/config
- https://raw.githubusercontent.com/void-linux/void-packages/master/srcpkgs/linux5.19/files/x86_64-dotconfig

Re: linux: tcp congestion and optimizations

[dtaht]

[-- Attachment #1: Type: text/plain, Size: 308 bytes --]

New comment by dtaht on void-packages repository

https://github.com/void-linux/void-packages/issues/39951#issuecomment-1332477148

Comment:
did you change the qdisc to fq_codel?

If boxes like this have sufficient horsepower, cake is actually the best thing going (IMHO, being one of the authors of both)