[Cake] openwrt was changing fq_codel quantum default

[Cake] openwrt was changing fq_codel quantum default

[Kevin Darbyshire-Bryant]

[-- Attachment #1: Type: text/plain, Size: 535 bytes --]

Did anyone else see https://dev.openwrt.org/ticket/21326 and fixing
commit https://dev.openwrt.org/changeset/47811 go through in the early
hours of this morning?



TLDR;

q->quantum = 300;  //was

q->quantum = psched_mtu(qdisc_dev(sch));   //now returned to kernel default

" i ran into cpu limits with sirq on archer c7v2 while testing ingress
shaping.

it seems to me that quantum < mtu is very costly (at cpu) with higher
(50mbps) bandwidths

    ~40% sirq load with quantum 1522+
    90% at quantum mtu/2"



[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4816 bytes --]

Re: [Cake] openwrt was changing fq_codel quantum default

[Dave Taht]

I did not like it when nbd changed the default quantum for the native qdisc.

So I support this change. That said, it has a negative effect at lower
rates (50mbit and below), when unshaped, and that bugs me also.
Dave Täht
Let's go make home routers and wifi faster! With better software!
https://www.gofundme.com/savewifi


On Tue, Dec 8, 2015 at 10:59 AM, Kevin Darbyshire-Bryant
<kevin@darbyshire-bryant.me.uk> wrote:
> Did anyone else see https://dev.openwrt.org/ticket/21326 and fixing
> commit https://dev.openwrt.org/changeset/47811 go through in the early
> hours of this morning?
>
>
>
> TLDR;
>
> q->quantum = 300;  //was
>
> q->quantum = psched_mtu(qdisc_dev(sch));   //now returned to kernel default
>
> " i ran into cpu limits with sirq on archer c7v2 while testing ingress
> shaping.
>
> it seems to me that quantum < mtu is very costly (at cpu) with higher
> (50mbps) bandwidths
>
>     ~40% sirq load with quantum 1522+
>     90% at quantum mtu/2"
>
>
>
> _______________________________________________
> Cake mailing list
> Cake@lists.bufferbloat.net
> https://lists.bufferbloat.net/listinfo/cake
>