On 12/11/2013 10:27 AM, Jan Lambertz wrote:
I found dd to be a very bad/misleading tool for this case.
Problems are caches in different layers of the system, filesystem
behaviour, sector sizing of drives and arrays, kernel configurations, input
data loading, real world scenarios and driver implementation.
I had same issues on centos.
Not perfect but a lot better for my purpose is bonnie++. Even with bonnie++
i would not dare to say that same tests on same hardware with centos and
openbsd will show the real differences in performance.
Maybe that might help to get more comparable results
Agreed. dd was a quick and dirty way to get some numbers after noticing
very unusual system performance with OpenBSD. I might have gotten a
little carried away with it. However, in this case I do think the
numbers generally correlate to my impression of overall disk performance
for this machine when running OpenBSD and FreeBSD. For example, when
unpacking the ports tree or compiling a kernel, FreeBSD seems to drive
the disks harder than OpenBSD (indicated by the drive activity lights,
drive noise, and the output scrolling across the screen). Of course,
this is hardly an objective metric and [for me] a ~15% disk I/O
performance difference is not terribly important. It is far more
important [to me] to have an elegant coherent reliable system with clean
source code and good documentation. If the underlying system is cobbled
together with zip-ties, duct-tape, and hot-rod hacks, it's not something
I could trust and I wouldn't invest too much in it. If the
zombie-apocalypse overruns the remnants of human society, it's the
OpenBSD source and documentation I want on my machines. ;)
