Re: Call for Adoption: TCP Tuning for HTTP

> On Mar 6, 2016, at 7:08 PM, Amos Jeffries <squid3@treenet.co.nz> wrote:
> 
> On 7/03/2016 2:17 p.m., Leif Hedstrom wrote:
>> 
>>> On Mar 1, 2016, at 10:46 PM, Mark Nottingham wrote:
>>> 
>>> [ copying Alison as our Transport Tech Advisor ]
>>> 
>>> Daniel has kindly started a document about how HTTP uses TCP, both for /1 and /2:
>>> <https://tools.ietf.org/html/draft-stenberg-httpbis-tcp>
>>> 
>>> We haven't explicitly discussed this at a meeting, but I have heard interest in this topic from a variety of folks.
>>> 
>>> What do people think about adopting this with a target of Best Current Practice?
>>> 
>>> Please comment on-list.
>> 
>> 
>> +1 on adopting this as a BCP.
>> 
>> I think its focus / bias towards Linux has to be addressed, which requires input / feedback from other OS vendors of course. But the adoption of this draft would likely help such efforts significantly.
>> 
>> Cheers,
>> 
>> — leif
> 
> 
> +1 "me too".
> 
> In regards to OS-specific things, I am somewhat against making it have
> any at all. The most desirable outcome of this draft would not be that
> admin can find what to tune for HTTP, but that TCP would evolve so we
> dont have to tune at all. Don't forget that HTTP applications require
> many different protocols all working together over the same transport(s)
> to be efficient. Tuning just for one wont help much.

Good point. I think “none" is better than "just Linux", which would force the authors to describe each option in a TCP and protocol standards way (and not implementation specific). Sounds like a lot of work though.

That much said, that might turn it into a difficult document to read. In the end, what many people need is a single sysctl.conf (etc.) file to deploy. :-).

Cheers,

— Leif

Received on Tuesday, 8 March 2016 22:48:48 UTC