Date: Mon, 7 May 2001 11:07:56 -0700 From: steve@Watt.COM (Steve Watt) To: pekkas@netcore.fi, stable@freebsd.org Subject: Re: 4.3-S: No buffer space available [SOLVED: dummynet] Message-ID: <200105071807.f47I7u731297@wattres.Watt.COM> References: <Pine.LNX.4.33.0105052014110.21251-100000@netcore.fi>
next in thread | previous in thread | raw e-mail | index | archive | help
pekkas@netcore.fi wrote: >On Sat, 5 May 2001, Pekka Savola wrote: >Ok. This was caused by the dummynet rule, either directly or indirectly. [...] >The traffic being shaped to 20Mbit/s ranged from 25-35 Mbit/s (steady), >mostly outgoing. >Has dummynet been tested in this kind of heavy environment? > >Is there a better value for 'queue', e.g. 1000Kbytes in this scenario? OK, this seems to me to be a guaranteed cause of "no buffer space available". If your pipe is only 20Mb and you're trying to stuff 25-35Mb though it, that's 5-15Mb that need to be stashed somewhere *every second*. A perfectly good place to stash that amount of stuff is /dev/null, unless you have some *huge* buffers allocated. Am I missing something? I don't know if dummynet supports it, but if you can enable random early dropping[1] (RED), that should do a decent job of throttling TCP connections to lower the demand. [1]OK, I *know* that's not what the RED acronym expands to, but it does the job... -- Steve Watt KD6GGD PP-ASEL-IA ICBM: 121W 56' 57.8" / 37N 20' 14.9" Internet: steve @ Watt.COM Whois: SW32 Free time? There's no such thing. It just comes in varying prices... To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200105071807.f47I7u731297>