xref: /dragonfly/share/man/man4/dummynet.4 (revision c89a6c1b)
1.\"
2.\" $FreeBSD: src/share/man/man4/dummynet.4,v 1.4.2.12 2002/11/18 21:51:16 luigi Exp $
3.\" $DragonFly: src/share/man/man4/dummynet.4,v 1.3 2007/10/16 11:28:40 sephe Exp $
4.\"
5.Dd October 28, 2002
6.Dt DUMMYNET 4
7.Os
8.Sh NAME
9.Nm dummynet
10.Nd traffic shaper, bandwidth manager and delay emulator
11.Sh DESCRIPTION
12.Em dummynet
13is a system facility that permits the control of traffic
14going through the various network interfaces, by applying bandwidth
15and queue size limitations, implementing different scheduling and queue
16management policies, and emulating delays and losses.
17.Pp
18The user interface for
19.Em dummynet
20is implemented by the
21.Nm ipfw
22program, so the reader is referred to the
23.Xr ipfw 8
24manpage for a complete description of the capabilities of
25.Nm
26and on how to use it.
27.Sh KERNEL OPTIONS
28The following options in the kernel configuration file are related to
29.Nm
30operation:
31.Bd -literal
32  IPFIREWALL               - enable ipfirewall (required for dummynet).
33  IPFIREWALL_VERBOSE       - enable firewall output.
34  IPFIREWALL_VERBOSE_LIMIT - limit firewall output.
35  DUMMYNET                 - enable dummynet operation.
36  NMBCLUSTERS              - set the amount of network packet buffers
37.Ed
38.Pp
39Generally, the following options are required:
40.Bd -literal
41  options IPFIREWALL
42  options DUMMYNET
43.Ed
44.Pp
45additionally, one may want to increase the number
46of mbuf clusters (used to store network packets) according to the
47sum of the bandwidth-delay products and queue sizes of all configured
48pipes.
49Timer granularity is controlled by
50.Xr sysctl 8
51variable
52.Va net.inet.ip.dummynet.hz ,
53which is set to 1000 by default.
54.Sh SEE ALSO
55.Xr setsockopt 2 ,
56.Xr bridge 4 ,
57.Xr ip 4 ,
58.Xr ipfw 8 ,
59.Xr sysctl 8
60.Sh HISTORY
61.Nm
62was initially implemented as a testing tool for TCP congestion control
63by
64.An Luigi Rizzo Aq luigi@iet.unipi.it ,
65as described on ACM Computer Communication Review, Jan.97 issue.
66Later it has been then modified to work at the ip and bridging
67level, integrated with the IPFW packet filter, and extended to
68support multiple queueing and scheduling policies.
69