Re: aio_read implementaion in Squid

Joe Cooper wrote:

> Yes, unfortunately. It seems none of the OS folks (even the free ones)
> want to see eye to eye on how to scale network event notification.
> FreeBSD has kqueues, Linux has about three thousand* different signals,
> event, and poll implementations in user and kernel space all competing
> to become official, and I'm not sure what the proprietary OS versions
> have other than the Solaris /dev/poll interface.

The Linux picture isn't that bad, but the main drawback seen is that
there is too much hacking and too little thought on I/O models there.
All recognise that I/O notifications is a big problem for large scale
servers, but instead of thinking the problem thru the first ideas that
pops up or is implemented in some other OS gets implemented in Linux..

> Thus the reason I think everyone (Henrik, Adrian and Robert, most of all
> I gather) wants to see Squid become a broker for different front and
> back end modules. i.e.

One of my goals at least. Squid should be able to utilize the best I/O
model of the platform where it runs, be this Linux, FreeBSD, Solaris,
Windows or whatever.

--
Henrik