Amos Jeffries schrieb:
Hi Amos,
> 5,000rps is around the fastest I have heard for Squid in production.
> 220mbit/s is about twice the bandwidth that user had. So well done on that.
Wow, didn't know that. But there is no magic included, just a plain
Squid. Good work on your side I would say.
> * turn zero_buffers off
> * experiment with the incoming_rate.
> * debugging in Squid is known to slow things down "debug_options
> ALL,0" is the fastest.
> * some small cache_mem on the CARP may or may not improve the overall
> speed. It will on the hottest objects.
I will try those and get back to the list.
> * tune the kernel TCP timeouts
This is a mystery to me. So I might get into this.
> If thats not fast enough for you, you could try the latest 2.HEAD and
> 3.HEAD for the latest speed boosters we have. 2.HEAD has the last of the
> Yahoo! and Adrian speed experiments and is usably stable. 3.HEAD is
> where we are working now, but may not be as fast as 2.x.
If my intention would be to break a new record I would try the HEAD
revision. But this thing should sooner or later go into a production
environment. So I better stick with the "official" stable version.
Cheers, Markus
Received on Thu Feb 04 2010 - 11:44:36 MST
This archive was generated by hypermail 2.2.0 : Thu Feb 04 2010 - 12:00:04 MST