Date: Fri, 23 Feb 2007 19:38:31 -0500 From: Charles Shannon Hendrix <shannon@widomaker.com> To: Doug Barton <dougb@FreeBSD.org> Cc: freebsd-stable@freebsd.org Subject: Re: getting garbage faster using FreeBSD? Message-ID: <20070223193831.3cc686c0@daydream.goid.lan> In-Reply-To: <45DF529B.9080101@FreeBSD.org> References: <45D9FD35.6040702@vwsoft.com> <20070219195143.GA42379@xor.obsecurity.org> <45DA121E.1040803@vwsoft.com> <20070220091238.c04cfceb.wmoran@collaborativefusion.com> <20070220165012.GB75535@xor.obsecurity.org> <20070220130051.bfcf6cf2.wmoran@collaborativefusion.com> <20070223113740.4da6e042@daydream.goid.lan> <45DF529B.9080101@FreeBSD.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, 23 Feb 2007 12:46:19 -0800
Doug Barton <dougb@FreeBSD.org> wrote:
> Charles Shannon Hendrix wrote:
>
> > Sometimes I'll get good speed for awhile (seconds), and then it starts
> > slowing down or stopping.
> >
> > /dev/random halts on me all the time.
> >
> > FreeBSD 6.1 on SPARC64
>
> First, usual advice of upgrading to the latest -stable, second, what
If you mean -STABLE, no thanks. This is a server, and I've never found
-STABLE to be stable.
> are the values of:
>
> sysctl kern.random
caesar# sysctl kern.random
kern.random.yarrow.gengateinterval: 10
kern.random.yarrow.bins: 10
kern.random.yarrow.fastthresh: 192
kern.random.yarrow.slowthresh: 256
kern.random.yarrow.slowoverthresh: 2
kern.random.sys.seeded: 1
kern.random.sys.harvest.ethernet: 1
kern.random.sys.harvest.point_to_point: 1
kern.random.sys.harvest.interrupt: 1
kern.random.sys.harvest.swi: 0
I manage to get a bursty 2.5MB/sec awhile ago. Bursty meaning periods of
reading and long pauses where reads block.
The hardware is a Sun Ultra 2, dual 300MHz CPUs.
2.5MB/sec would be fine for most of my uses if it were steady. The problem
is the long pauses where I get 0MB/sec.
For example, when I'm streaming an encrypted backup over the net from this
machine, it would be nice if /dev/random I/O didn't cause any pauses.
Most of the time it doesn't affect much, but it seems like this machine
shouldn't have such a slow /dev/random, and I thought the new code didn't
block in any case.
--
shannon | The trade of governing has always been monopolized by the
| most ignorant and the most rascally individuals of
| mankind.
| -- Thomas Paine
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20070223193831.3cc686c0>