Date: 06 Jul 2003 11:45:08 -0400 From: Lowell Gilbert <freebsd-stable-local@be-well.no-ip.com> To: stable@freebsd.org Subject: Re: Weird vmstat -s stats Message-ID: <44wuevveuz.fsf@be-well.ilk.org> In-Reply-To: <200307060029.00866.me@farid-hajji.de> References: <200307051728.24681.me@farid-hajji.de> <44brw8g26e.fsf@be-well.ilk.org> <200307060029.00866.me@farid-hajji.de>
index | next in thread | previous in thread | raw e-mail
Farid Hajji <me@farid-hajji.de> writes: > > > "vmstat -s" shows on one of my boxes (uptime: 26 days): > > > > > > -1597015721 total name lookups > > > cache hits (101% pos + 0% neg) system 0% per-directory > > > deletions 0%, falsehits 0%, toolong 0% > > > > > > Weird. I'll have to cvsup again and hope this has been fixed :) > > > > I'm not seeing it on an up-to-date -STABLE. Of course, that doesn't > > mean much, because it looks like a counter wraparound. > > Of course it's a wraparound. The total count is a wraparound, but the positive cache hit count still looks funny. > I was not able to reproduce this > condition on newer -STABLE. It takes some weeks of uptime > and heavy use to reach this stage, unless you happen to know > how to increment the counter with a test program (I don't > know how). You'd have to put in an extra kernel interface for that; the ability to zero the structure would be (slightly) more useful. > Shouldn't such counters be at least 64 bit wide? They are, on most platforms. You can't do that on an i386, though. At least, not without opening up worse problems.home | help
Want to link to this message? Use this
URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?44wuevveuz.fsf>