Re: [squid-users] Re: Accelerating Proxy options?

 On Tue, 19 Apr 2011 19:11:31 -0700, Linda Walsh wrote:
> Amos Jeffries wrote:
>> On Tue, 19 Apr 2011 13:31:38 -0700, Linda Walsh wrote:
>>> Picture this: I (on a client sys) pull in a web page. At same
>>> time
>>> I get it, it's handed over to a separate process running on a
>>> separate core
>> read -> copy to reader thread buffer -> copy to processing thread
>> buffer -> copy to result output buffer (maybe) -> coy to writer thread
>> buffer -> write.
>> 2x slowdown *on top of* the above processing scan lags. This exact
>> case of multiple copying is one of two reasons we do not have
>> threading in Squid.
> ----
> I think you are missing an important point (maybe?)
> The current fetch would stay as is. I.e go str8 to client --
> only after sent to client would a copy be sent off to the
> 'prefetcher', so at worst, prefetcher does it's scan and finds out
> stuff is already
> fetched by client-requests, but in best case, prefetcher would do
> it's parsing before the client asked for the page-reqs and they
> would,
> theoretically, already be 'in process' of being fetched.

 Ah. I see.

>
>>> Anyway, just some wonderings...
>>> What will it take for Sq3 to get to the feature level of Sq2
>>> and allow,
>> What we are missing in a big way is store-URL and location-URL
>> re-writers.
> ===
> I'm so out of it, I don't even know that terminology -- I could
> 'guess', like maybe 1) being rewriting the URL of content stored in
> the cache to
> be more easily reusable, and 2) same as 1, but hmmm...rewriting
> requests from clients?... Ah heck...I'm probably writing out my toes
> again...
> (like speaking out some wrong orifice, but for writing v. speaking!
> ;-)).

 You may be more familiar with the config directive names:
 storeurl_rewrite_program and location_rewrite_program.
 storeurl_* is the feature used to cache YouTube videos.

>
> Oh well nevermind...my floor is covered with drippings from
> overflowing plate.
> **slog**
>
>
>> I've had a handful of people stick their hands up to do this over
>> the last year or two. Pointed them at the squid-2 patches which need
>> adjusting to compile and work in squid-3 code. Never to hear from them
>> again. :(
> ---
> Sigh...probably mean well, but when they get in there, they find
> they are way over their heads...or at least alot more work than they
> thought it'd be... not that I don't run into that many times in
> trying
> to change something or other on OpnSrc sw...geez...
> Hardly anything is just simple anymore...
> My home network is held together with bailing wire and duct tape
> and even then it's unstable in weird ways.

 lol.

 Amos
Received on Wed Apr 20 2011 - 02:39:35 MDT