Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 11 May 95 21:01:19 MDT
From:      terry@cs.weber.edu (Terry Lambert)
To:        freebsd-hackers@FreeBSD.org
Cc:        taob@gate.sinica.edu.tw
Subject:   Re: Apache + FreeBSD 2.0 benchmark results
Message-ID:  <9505120301.AA12192@cs.weber.edu>
In-Reply-To: <199505112230.BAA29734@lk-hp-20.hut.fi> from "Juha Inkari" at May 12, 95 01:30:30 am

next in thread | previous in thread | raw e-mail | index | archive | help
> > from.  What you have is an I/O Dispatching server according to what I've
> > been taught .
> 
> I guess the point is differentiating servers in how they perform
> multi-tasking. The server could dispatch several requests to different
> tasks or it could process the requests syncronously consuming several
> tasks for one request. Here goes yet another set of names:

I/O dispatch and event dispatch models are not necessarily multitasking
at all; they run in a flat stack space, so they aren't even really tasks.

> - Repeated forking heavyweight processing server.
> - Multiple forked heavyweight processing server.
> - Lightweight processing server.
> 
> Also, multi-threaded or asyncronous could be added for clarification
> in some contexts. The "lightweight processing server" says nothing
> about the time of creation of the processing tasks, or about the
> "lightness" of the thread of control, except that it should not be as
> resource consuming that the "heavyweight" model.

This is getting rather silly, and is devolving quickly into fictional
semantics.

I have an exhaustive list of the tasking models I considered when working
with Drew Spencer on the NWU 4.0 code which I will send (not to the list)
under seperate cover when I get it done.  You may have to kick me in a
week and again a week after that to get me to hop to it, though.  It lists
the models, the implementations available for each model type, and the
tradeoffs they entail.

Unfortunately, there's not a good Knuth book I can point you at (we all
know why that is).


					Terry Lambert
					terry@cs.weber.edu
---
Any opinions in this posting are my own and not those of my present
or previous employers.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9505120301.AA12192>