The lattest article is a presentation in 2006 and unfortunately it's not as detailed (since it's a presentation) as the 2003 article. The perfomance data measured in 2003 might vary as Bernard says given current date OS and hardware. Atleast the code seems available. I might run some test when I get a couple of free weeks. If I do, I'll post the results on the list (and start a flame war HAHAHAHA)
8~D
--- On Tue, 9/29/09, Bernard Owuor <b_owuor@yahoo.com> wrote:
From: Bernard Owuor <b_owuor@yahoo.com>
Subject: Re: [Skunkworks] scalability benchmarks
To: "Skunkworks Forum" <skunkworks@lists.my.co.ke>
Date: Tuesday, September 29, 2009, 2:30 AM
I wish I had some time to run these tests on a FreeBSD 7.0 and above. The freebsd appeared to oublaze the 2.6.21 kernel when I benchmarked some time last year - especially on the networking side. I suspected that the file-system was faster too - but someone oughtta check this out before swallowing the data. Also, the tests were done on a single CPU machine, there could be marked performance differences on smp machines.
From: saidimu apale <saidimu@gmail.com>
To: Skunkworks Forum <skunkworks@lists.my.co.ke>
Sent: Monday, September 28, 2009 4:54:35 PM
Subject: Re: [Skunkworks] scalability benchmarks
Wesley, what was your biggest surprise?
Mine was the perfomance of journaled vs. non-journaled filesystems. It was a rather pleasant surprise. One of the biggest benefits, to me, of *nix is the variety of filesystems and the ability to mix and match the filesystems for maximum performance, e.g. one partition formatted with an fs that does well with many small files (but performs poorly for a few massive files) and another parition with an fs that does great with relatively few massive files (e.g. media files) but does badly with a huge number of small files (e.g. temp files from a webserver cache).
saidi
On Mon, Sep 28, 2009 at 11:48 AM, wesley kirinya <kiriinya2000@yahoo.com> wrote:
Big thanx for this one!
--- On Sat, 9/26/09, saidimu apale <saidimu@gmail.com> wrote:
From: saidimu apale <saidimu@gmail.com>
Subject: [Skunkworks] scalability benchmarks
To: "Skunkworks forum" <skunkworks@lists.my.co.ke>
Date: Saturday, September 26, 2009, 7:40 AM
http://bulk.fefe.de/An informative, and fairly technical, read on network and filesystem scalability bechmarks. The first 2-links are Unix-only (*BSD and Linux), the last link includes various Windows and Solaris versions. What comes out on top? The answer won't surprise you, neither will the identity of the worst-performing OS. What is shocking is how badly some otherwise decent OSes perform. The section on filesystems is also quite interesting: you thought non-journalled filesytems were faster than journalled ones? Think again. Quick FAQ on the article: O(n) means whatever is being measured proportionally increases as n increases, O(1) means performance is constant no matter how big n is. For instance: if you were measuring the amount of time it takes a webserver to serve requests, an O(n) performance means the time the webserver needs to serve a
request increases proportionally to the number of requests; an O(1) performance means it takes the same amount of time for the webserver to serve 1 request as it would a gazillion bazillion requests. saidi -----Inline Attachment Follows-----
|
-----Inline Attachment Follows-----
|