Date: Tue, 17 May 2005 00:27:53 +0400 From: Gleb Smirnoff <glebius@FreeBSD.org> To: Donatas <donatas@lrtc.net> Cc: freebsd-net@FreeBSD.org Subject: Re: flood with ng_bridge Message-ID: <20050516202753.GB86723@cell.sick.ru> In-Reply-To: <03c301c55a55$0145d180$9f90a8c0@DONATAS> References: <008701c559da$a9e1b050$9f90a8c0@DONATAS> <20050516130820.GA83360@cell.sick.ru> <03c301c55a55$0145d180$9f90a8c0@DONATAS>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, May 16, 2005 at 11:22:38PM +0300, Donatas wrote: D> > D> 3. Monitorring traffic with mrtg on ngethxxx and hatm0 interfaces we can se interesting output amplitude fluctations: D> > D> ftp://temp:temp@217.9.241.242/hatm0.png D> > D> seems that origin of those fluctations is ng_bridge. D> D> > What is the timescale? How long one peak/pit lasts in time? I doubt that this D> > is related to ng_bridge. D> crontab: D> ### Regular MRTG monitoring D> 0,10,20,30,40,50 * * * * root /usr/local/etc/mrtg/exec_mrtg.sh > /dev/null 2>&1 No-no. The question was: what is the scale of time axis? How long does one peak last? D> we have ng_bridge, (bridged: fxp0, ngeth(1-xxx), hatm0) D> D> ngethxxx are linked with hatm0 in the following way: D> hatm0<->atmllc<->ng_vlan<->ng_eiface<-> Can you please show the script you use to build graph? D> we've tried tens of variuos configurations on determining the couse of those fluctuations. They seems to be not only in ng_bridge but in kernel.bridge as well. We've tried to change all 3 parameters in ng_bridge and still - those parameters doesn't seems to affect fluctuations. Have you doublechecked that fluctuation aren't caused just by the source of traffic? For example some host runs periodic job... -- Totus tuus, Glebius. GLEBIUS-RIPN GLEB-RIPE
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050516202753.GB86723>