Date: Tue, 24 Jul 2001 13:30:55 -0400 (EDT) From: Joe Clarke <marcus@marcuscom.com> To: Chris Phillips <chris@selkie.org> Cc: <freebsd-questions@FreeBSD.ORG> Subject: Re: FreeBSD Router Message-ID: <20010724132536.G6609-100000@shumai.marcuscom.com> In-Reply-To: <Pine.BSF.4.21.0107241001530.262-100000@shell.bchosting.com>
next in thread | previous in thread | raw e-mail | index | archive | help
I t could be your network cards. On my production "router" I have two Intel EtherExpress Pros. On my home router, I have a ed0 and a EtherExpress Pro. I also have a much shorter ipfw ruleset (though I'm not sure that could be causing this). I know the cheap Linksys 10/100 cards cause me some problems. Joe Clarke On Tue, 24 Jul 2001, Chris Phillips wrote: > Hey Joe, > > [root@gateway1 /root]# netstat -m > 489/784/34816 mbufs in use (current/peak/max): > 488 mbufs allocated to data > 1 mbufs allocated to packet headers > 332/528/8704 mbuf clusters in use (current/peak/max) > 1252 Kbytes allocated to network (62% in use) > 0 requests for memory denied > 0 requests for memory delayed > 0 calls to protocol drain routines > > Everything appears to be fine there unless I am not reading the output > right. > > Any other suggestions? > > Thanks. > > -Chris Phillips > > On Tue, 24 Jul 2001, Joe Clarke wrote: > > > Usually packet loss starts to occur when buffer space runs short (or I/O > > mem on Cisco routers). How many MBUFs do you have configured? What does > > netstat -m say? You may want to tune your MBUFs in the kernel. IF you > > have a current CVS of -stable, checkout the tuning(7) manpage. The thing > > to tune would be NMBCLUSTERS. For my large machines, I set it to 16384. > > The default value is 1024, I think. > > > > Joe Clarke > > > > On Tue, 24 Jul 2001, Chris Phillips wrote: > > > > > Hey, > > > > > > I always hear people joking about how their 486 would make a great > > > router. I have a P166 w/ 128MB ram and three intel 10/100 NICs running as > > > my firewall/router. During peak times it is starting to have packet > > > loss. At what point is it too much for this kind of hardware? Does > > > anyone have any similar experiences? Is the packet loss a direct result > > > >from lack of processing power? There is approximately 250 ipfw rules. > > > > > > [root@gateway1 /root]# w > > > 9:26AM up 414 days, 12:54, 1 user, load averages: 0.00, 0.00, 0.00 > > > USER TTY FROM LOGIN@ IDLE WHAT > > > root p0 penguin Thu08PM - w > > > > > > [root@gateway1 /root]# vmstat > > > procs memory page disks faults cpu > > > r b w avm fre flt re pi po fr sr ad0 md0 in sy cs us sy > > > id > > > 0 0 0 3236 79424 1 0 0 0 1 0 0 0 771 7 2 0 7 > > > 93 > > > > > > Anyone have any other suggestions for the packet loss? > > > > > > Peak times can be up to 6mbit/s sustained. This box routes 3 different > > > subnets that add up to 256 IPs. There are two Catalyst 3524s sitting > > > behind it. I'm not sure what other info I should include. > > > > > > All comments/suggestions are appreciated. Thanks. > > > > > > -Chris Phillips > > > > > > P.S. Please CC me as I am not on the list. > > > > > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org > > > with "unsubscribe freebsd-questions" in the body of the message > > > > > > > > > > > > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010724132536.G6609-100000>