Date: Thu, 19 Dec 2002 06:58:31 +0700 From: "Hendry S." <hsa@zeus.havoc2k.or.id> To: Takashi Okumura <taka@wide.ad.jp> Cc: freebsd-net <freebsd-net@FreeBSD.ORG> Subject: Re: [Netnice version 2 (For FreeBSD 4.7)] Message-ID: <20021218235831.GA22748@zeus.havoc2k.or.id> In-Reply-To: <3DFF57F1.AA2E7CD9@wide.ad.jp> References: <3DFF57F1.AA2E7CD9@wide.ad.jp>
next in thread | previous in thread | raw e-mail | index | archive | help
Once upon a time, Takashi Okumura said that > Dear Hackers, > > i would like to introduce my traffic control package for FreeBSD 4.7, > netnice version 2, and recruit some alpha testers. > > > ============================================================ > Netnice Version 2: Alpha Release (Dec 10, 2002) > ============================================================ > > > Netnice is a traffic management package for FreeBSD. > It comes with useful traffic control applications; > > o Bandwidth control of running network applications: netnice > > % ps -aux | grep netscape > taka 1234 ... snip ... 9:14AM 0:00.10 netscape > % netnice 1234 1024Kbps > > o Easy specification of network I/O at startup: sh > > $ ftp ftp.freebsd.org @512Kbps > > o Bandwidth specification in /etc/inetd.conf: inetd > > % cat /etc/inetd.conf > ftp stream nowait ...snip... @ 512Kbps > telnetd stream nowait ...snip... @ 64Kbps > > o Network QoS daemon with flexible scripting feature: netniced > > % netniced -d > > > * Overview * > > In FreeBSD community, we already have some implementations for traffic > control. Most of them have parallel queues in network subsystem of the > operating system, and basically, follow traditional traffic control > model where we configure queue parameters and packet classification > rules. > > Netnice differs from them in that it utilizes a new traffic control > primitive, Virtual network interface (VIF). Traffic control with a VIF > works as follows. First, we create a VIF connected to an existing > physical network interface. Then, we set parameters of the VIF, such > as bandwidth and priority. Lastly, we attach it to network I/O of > target process, or sockets. > > > * Features * > > The primitive has following notable features. > > The attached virtual network interface will be inherited to its child > processes, and thus, the network resource would be protected from > excess use. > > Virtual network interfaces can be hierarchically structured, and > attachable to sockets and process groups, as well as to process > network I/O. This model provides flexible control granularity. > > The virtual network interface has various parameters, such as > bandwidth, queuing discipline, weight and priority. Hence, users can > enjoy various traffic control service, such as software packet shaping > and premium service, in a single framework. > > Lastly, we note that the virtual network interfaces are accessible as > files, under /proc/network, and thus, any programming environment with > file access API can make use of the primitive. > > > * Needs alpha testers!! * > > If you got interested in the implementation, please join the alpha > testing, working towards beta-version release in near future. If there > were requests for for an Apache module to control Web traffic, i think > i can provide an experimental version shortly. Detailed information, > documents, and the package is available at the project homepage. > > Project Homepage http://www.netnice.org > > > thanks! wireless% host www.netnice.org Host www.netnice.org not found: 2(SERVFAIL) cheers Hendry S. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021218235831.GA22748>