Date: Tue, 3 Jan 2006 00:42:23 -0800 From: Gary Kline <kline@tao.thought.org> To: James Long <list@museum.rain.com> Cc: Gary Kline <kline@tao.thought.org>, freebsd-questions@freebsd.org Subject: Re: STressing a new server... Message-ID: <20060103084223.GA43443@thought.org> In-Reply-To: <20060102180652.GA81087@ns.museum.rain.com> References: <20060102105418.EA25316A433@hub.freebsd.org> <20060102180652.GA81087@ns.museum.rain.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Jan 02, 2006 at 10:06:52AM -0800, James Long wrote: > A buildworld is indeed an excellent test of memory, CPU, drives > and cabling. > > The memory tester is sysutils/memtest. > > It's not a very scientific test, but one thing I do is a > make -j8 buildworld whilst I do a large tar operation, > optionally with compression. Just anecdotally it appears > that tar likes to use lots of memory. I usually do some > sort of file system move, ala: > > tar jcf - original | tar xpvf - -C copy > > Make sure you have enough disk space to burn. > > This has caught dodgy memory on servers in the past. > > Thanks for the tar pointer. Got to copy over /home here to the new box and run several into /usr/tmp. Thanks to everybody for ideas. I don't want to stress things until something *melts*... but some reasonably hard use. I'm using three or four benchmarks, and a couple of my own tests. xload and top show that the results are a fair simulation of real-world use. Should know by the 22nd! gary -- Gary Kline kline@thought.org www.thought.org Public service Unix
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060103084223.GA43443>