Date: Sat, 21 Mar 2009 20:14:08 +0300 From: Sergey Matveyhcuk <sem@FreeBSD.org> To: Eugene L Kovalenja <qwe@qwe.net.ua> Cc: freebsd-ipfw@freebsd.org Subject: Re: FreeBSD 7.0: dummynet 99% cpu Message-ID: <49C52060.6060504@FreeBSD.org> In-Reply-To: <49C4F3EE.2080903@qwe.net.ua> References: <49C04CA3.1070100@qwe.net.ua> <49C35EB3.6040508@FreeBSD.org> <49C4F3EE.2080903@qwe.net.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
Well, I did not test the patch. Quick idea was to decrease hash buckets increasing of hash size. I did not dig deeper dummynet. Eugene L Kovalenja wrote: > Sergey Matveyhcuk пишет: >> Could you test an included patch? >> >> Eugene L Kovalenja wrote: >>> >>> Time in three days traffic via ipfw doesn't go. In top: >>> 21 root 1 -44 - 0K 8K WAIT 7 2:15 99.02% >>> dummynet >>> (this is example, not copy\paste) >>> >>> Also sw1: net increases from 5-10% to 30-35%... >>> >>> I am helped only by reboot. >>> >>> In what can consist the problem? >>> >> >> ------------------------------------------------------------------------ >> >> _______________________________________________ >> freebsd-ipfw@freebsd.org mailing list >> http://lists.freebsd.org/mailman/listinfo/freebsd-ipfw >> To unsubscribe, send any mail to "freebsd-ipfw-unsubscribe@freebsd.org" > Hello. > > Thank you for this patch. > > After kernel patching trouble changes: > Mar 21 13:58:45 taurus kernel: bge0: watchdog timeout -- resetting > Mar 21 13:58:45 taurus kernel: bge0: link state changed to DOWN > > In top utility 99% cpu on irq bge0. > > Changing of network adapter (on board 2 adapters: bge0 && bge1) didn't > decide problem. > > my /boot/loader.conf > [root@taurus /usr/home/login]# cat /boot/loader.conf > kern.ipc.maxpipekva=67108864 > net.graph.maxalloc=2048 > > #vm.kmem_size_max="512M" > #vm.kmem_size="512M" > kern.maxusers="512" > > :(
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?49C52060.6060504>