Date: Sat, 17 Feb 2007 17:08:00 +0100 From: Alexander Leidinger <Alexander@Leidinger.net> To: "Steven H. Baeighkley" <stevenb@frii.com> Cc: freebsd-performance@freebsd.org Subject: Re: bizarre performance issues in 6.2 release Message-ID: <20070217170800.l0au44ygowkw0s0w@webmail.leidinger.net> In-Reply-To: <45D640FE.3070805@frii.com> References: <45D640FE.3070805@frii.com>
next in thread | previous in thread | raw e-mail | index | archive | help
Quoting "Steven H. Baeighkley" <stevenb@frii.com> (from Fri, 16 Feb =20 2007 16:40:46 -0700): > These servers are running apache, perl, proftpd and php. One server is > running 4.11 and is rocking the world, load is rarely above 1 and is > regularly below .5. The other server is running 6.2 and regularly has > load between 2 and 6 despite serving fewer connections than the 4.11 > box. Using the load of unequal operating system versions as a measure for =20 performance is not a good idea. In fact, the algorithm to count the =20 load did change between 4.x and 5.x (the numbers are larger now, but =20 the box should be able to do the same amount of work). A more =20 appropriate value to measure the performance is to benchmark the =20 number of handled connections to a service in a specific timeframe, or =20 the min/max/avg response time of a request. I suggest to have a look =20 at "ab", "jmeter" or similar tools to measure the performance of a web =20 server. When those numbers are different by an significant amount, feel free =20 to come back and ask for suggestions (but first google around please, =20 accf_http and libthr may make a difference and are well documented). Bye, Alexander. --=20 Always leave room to add an explanation if it doesn't work out. http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID =3D B0063FE7 http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID =3D 72077137
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070217170800.l0au44ygowkw0s0w>