Discussion:
Update Cake page on bufferbloat.net?
(too old to reply)
Rich Brown
2017-11-03 14:53:57 UTC
Permalink
Raw Message
Hi folks,

I saw a blog posting that was enthusing about codel/fq_codel, and I was moved to respond that the state of the art was now cake.

But I looked at the Cake page on Bufferbloat.net and wonder if everything there is true, or whether it would be good to update it. https://www.bufferbloat.net/projects/codel/wiki/Cake/

(Once the technical information on that page is current, I would also be willing to tackle a modest reorganization to add some background info to the beginning so that we can cite that page as the primary reference for Cake.)

Thanks!

Rich
Dave Taht
2017-11-03 18:59:04 UTC
Permalink
Raw Message
Post by Rich Brown
Hi folks,
I saw a blog posting that was enthusing about codel/fq_codel, and I was moved to
respond that the state of the art was now cake.
where?
Post by Rich Brown
But I looked at the Cake page on Bufferbloat.net and wonder if everything there
is true, or whether it would be good to update
it. https://www.bufferbloat.net/projects/codel/wiki/Cake/
It's pretty much true.
Post by Rich Brown
(Once the technical information on that page is current, I would also be willing
to tackle a modest reorganization to add some background info to the beginning
so that we can cite that page as the primary reference for Cake.)
I would really for us to make the effort to upstream this in 2018. I'm
still dubious about "cobalt", and I'd like to make the GRO peeling less
agressive (again), relative to load, and it needs some cleanups to match
current kernels.
Post by Rich Brown
Thanks!
Rich
_______________________________________________
Bloat mailing list
https://lists.bufferbloat.net/listinfo/bloat
Rich Brown
2017-11-03 19:28:19 UTC
Permalink
Raw Message
Post by Dave Taht
Post by Rich Brown
I saw a blog posting that was enthusing about codel/fq_codel, and I was moved to
respond that the state of the art was now cake.
where?
This article (https://www.pcmech.com/article/bufferbloat-fix-slow-network/ <https://www.pcmech.com/article/bufferbloat-fix-slow-network/>) is a pretty sub-standard explanation of bufferbloat. But I didn't want to come across as the "smartest (smart-ass) guy in the room"

My plan was to gently correct the worst errors/misperceptions ("it's the bottleneck, stupid") and say that the state of the art had moved ahead, even of fq_codel, then point to the Cake page on the site.
Post by Dave Taht
Post by Rich Brown
But I looked at the Cake page on Bufferbloat.net <http://bufferbloat.net/> and wonder if everything there
is true, or whether it would be good to update
it. https://www.bufferbloat.net/projects/codel/wiki/Cake/ <https://www.bufferbloat.net/projects/codel/wiki/Cake/>
It's pretty much true.
Good. I will try this weekend to organize that info into a page that would serve well readers like those from the pcmech.com <http://pcmech.com/> site who're new to the subject, and curious about Cake/Bufferbloat.

Rich
Jonathan Morton
2017-11-03 19:35:22 UTC
Permalink
Raw Message
I think "mostly true but incomplete" is the best way to describe the
current pages.

- Jonathan Morton
Toke Høiland-Jørgensen
2017-11-03 19:39:48 UTC
Permalink
Raw Message
Post by Jonathan Morton
I think "mostly true but incomplete" is the best way to describe the
current pages.
Patches welcome :)

https://github.com/tohojo/bufferbloat-net

-Toke

Loading...