man/man4/multicast.4

.\" Copyright (c) 2001-2003 International Computer Science Institute
.\"
.\" Permission is hereby granted, free of charge, to any person obtaining a
.\" copy of this software and associated documentation files (the "Software"),
.\" to deal in the Software without restriction, including without limitation
.\" the rights to use, copy, modify, merge, publish, distribute, sublicense,
.\" and/or sell copies of the Software, and to permit persons to whom the
.\" Software is furnished to do so, subject to the following conditions:
.\"
.\" The above copyright notice and this permission notice shall be included in
.\" all copies or substantial portions of the Software.
.\"
.\" The names and trademarks of copyright holders may not be used in
.\" advertising or publicity pertaining to the software without specific
.\" prior permission. Title to copyright in this software and any associated
.\" documentation will at all times remain with the copyright holders.
.\"
.\" THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
.\" IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
.\" FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
.\" AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
.\" LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
.\" FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
.\" DEALINGS IN THE SOFTWARE.
.\"
.\" $FreeBSD: src/share/man/man4/multicast.4,v 1.4 2004/07/09 09:22:36 ru Exp $
.\" $OpenBSD: multicast.4,v 1.8 2013/08/14 08:46:07 jmc Exp $
.\" $NetBSD: multicast.4,v 1.3 2004/09/12 13:12:26 wiz Exp $
.\"
.Dd $Mdocdate: August 14 2013 $
.Dt MULTICAST 4
.Os
.\"
.Sh NAME
.Nm multicast
.Nd Multicast Routing
.\"
.Sh SYNOPSIS
.Cd "options MROUTING"
.Pp
.In sys/types.h
.In sys/socket.h
.In netinet/in.h
.In netinet/ip_mroute.h
.In netinet6/ip6_mroute.h
.Ft int
.Fn getsockopt "int s" IPPROTO_IP MRT_INIT "void *optval" "socklen_t *optlen"
.Ft int
.Fn setsockopt "int s" IPPROTO_IP MRT_INIT "const void *optval" "socklen_t optlen"
.Ft int
.Fn getsockopt "int s" IPPROTO_IPV6 MRT6_INIT "void *optval" "socklen_t *optlen"
.Ft int
.Fn setsockopt "int s" IPPROTO_IPV6 MRT6_INIT "const void *optval" "socklen_t optlen"
.Sh DESCRIPTION
.Tn "Multicast routing"
is used to efficiently propagate data
packets to a set of multicast listeners in multipoint networks.
If unicast is used to replicate the data to all listeners,
then some of the network links may carry multiple copies of the same
data packets.
With multicast routing, the overhead is reduced to one copy
(at most) per network link.
.Pp
All multicast-capable routers must run a common multicast routing
protocol.
The Distance Vector Multicast Routing Protocol (DVMRP)
was the first developed multicast routing protocol.
Later, other protocols such as Multicast Extensions to OSPF (MOSPF),
Core Based Trees (CBT),
Protocol Independent Multicast \- Sparse Mode (PIM-SM),
and Protocol Independent Multicast \- Dense Mode (PIM-DM)
were developed as well.
.Pp
To start multicast routing,
the user must enable multicast forwarding via the
.Xr sysctl 8
variables
.Va net.inet.ip.mforwarding
and/or
.Va net.inet.ip6.mforwarding .
The user must also run a multicast routing capable user-level process,
such as
.Xr mrouted 8 .
From a developer's point of view,
the programming guide described in the
.Sx Programming Guide
section should be used to control the multicast forwarding in the kernel.
.\"
.Ss Programming Guide
This section provides information about the basic multicast routing API.
The so-called
.Dq advanced multicast API
is described in the
.Sx "Advanced Multicast API Programming Guide"
section.
.Pp
First, a multicast routing socket must be open.
That socket would be used
to control the multicast forwarding in the kernel.
Note that most operations below require certain privilege
(i.e., root privilege):
.Bd -literal -offset indent
/* IPv4 */
int mrouter_s4;
mrouter_s4 = socket(AF_INET, SOCK_RAW, IPPROTO_IGMP);
.Ed
.Bd -literal -offset indent
int mrouter_s6;
mrouter_s6 = socket(AF_INET6, SOCK_RAW, IPPROTO_ICMPV6);
.Ed
.Pp
Note that if the router needs to open an IGMP or ICMPv6 socket
(IPv4 or IPv6, respectively)
for sending or receiving of IGMP or MLD multicast group membership messages,
then the same
.Va mrouter_s4
or
.Va mrouter_s6
sockets should be used
for sending and receiving respectively IGMP or MLD messages.
In the case of
.Bx -derived
kernels,
it may be possible to open separate sockets
for IGMP or MLD messages only.
However, some other kernels (e.g.,
.Tn Linux )
require that the multicast
routing socket must be used for sending and receiving of IGMP or MLD
messages.
Therefore, for portability reasons, the multicast
routing socket should be reused for IGMP and MLD messages as well.
.Pp
After the multicast routing socket is open, it can be used to enable
or disable multicast forwarding in the kernel:
.Bd -literal -offset 5n
/* IPv4 */
int v = 1;        /* 1 to enable, or 0 to disable */
setsockopt(mrouter_s4, IPPROTO_IP, MRT_INIT, (void *)&v, sizeof(v));
.Ed
.Bd -literal -offset 5n
/* IPv6 */
int v = 1;        /* 1 to enable, or 0 to disable */
setsockopt(mrouter_s6, IPPROTO_IPV6, MRT6_INIT, (void *)&v, sizeof(v));
\&...
/* If necessary, filter all ICMPv6 messages */
struct icmp6_filter filter;
ICMP6_FILTER_SETBLOCKALL(&filter);
setsockopt(mrouter_s6, IPPROTO_ICMPV6, ICMP6_FILTER, (void *)&filter,
           sizeof(filter));
.Ed
.Pp
After multicast forwarding is enabled, the multicast routing socket
can be used to enable PIM processing in the kernel if either PIM-SM or
PIM-DM are being used
(see
.Xr pim 4 ) .
.Pp
For each network interface (e.g., physical or a virtual tunnel)
that would be used for multicast forwarding, a corresponding
multicast interface must be added to the kernel:
.Bd -literal -offset 3n
/* IPv4 */
struct vifctl vc;
memset(&vc, 0, sizeof(vc));
/* Assign all vifctl fields as appropriate */
vc.vifc_vifi = vif_index;
vc.vifc_flags = vif_flags;
vc.vifc_threshold = min_ttl_threshold;
vc.vifc_rate_limit = max_rate_limit;
memcpy(&vc.vifc_lcl_addr, &vif_local_address, sizeof(vc.vifc_lcl_addr));
if (vc.vifc_flags & VIFF_TUNNEL)
    memcpy(&vc.vifc_rmt_addr, &vif_remote_address,
           sizeof(vc.vifc_rmt_addr));
setsockopt(mrouter_s4, IPPROTO_IP, MRT_ADD_VIF, (void *)&vc,
           sizeof(vc));
.Ed
.Pp
The
.Va vif_index
must be unique per vif.
The
.Va vif_flags
contains the
.Dv VIFF_*
flags as defined in
.Aq Pa netinet/ip_mroute.h .
The
.Va min_ttl_threshold
contains the minimum TTL a multicast data packet must have to be
forwarded on that vif.
Typically, it would be 1.
The
.Va max_rate_limit
contains the maximum rate (in bits/s) of the multicast data packets forwarded
on that vif.
A value of 0 means no limit.
The
.Va vif_local_address
contains the local IP address of the corresponding local interface.
The
.Va vif_remote_address
contains the remote IP address for DVMRP multicast tunnels.
.Bd -literal -offset indent
/* IPv6 */
struct mif6ctl mc;
memset(&mc, 0, sizeof(mc));
/* Assign all mif6ctl fields as appropriate */
mc.mif6c_mifi = mif_index;
mc.mif6c_flags = mif_flags;
mc.mif6c_pifi = pif_index;
setsockopt(mrouter_s6, IPPROTO_IPV6, MRT6_ADD_MIF, (void *)&mc,
           sizeof(mc));
.Ed
.Pp
The
.Va mif_index
must be unique per vif.
The
.Va mif_flags
contains the
.Dv MIFF_*
flags as defined in
.Aq Pa netinet6/ip6_mroute.h .
The
.Va pif_index
is the physical interface index of the corresponding local interface.
.Pp
A multicast interface is deleted by:
.Bd -literal -offset indent
/* IPv4 */
vifi_t vifi = vif_index;
setsockopt(mrouter_s4, IPPROTO_IP, MRT_DEL_VIF, (void *)&vifi,
           sizeof(vifi));
.Ed
.Bd -literal -offset indent
/* IPv6 */
mifi_t mifi = mif_index;
setsockopt(mrouter_s6, IPPROTO_IPV6, MRT6_DEL_MIF, (void *)&mifi,
           sizeof(mifi));
.Ed
.Pp
After multicast forwarding is enabled, and the multicast virtual
interfaces have been
added, the kernel may deliver upcall messages (also called signals
later in this text) on the multicast routing socket that was open
earlier with
.Dv MRT_INIT
or
.Dv MRT6_INIT .
The IPv4 upcalls have a
.Vt "struct igmpmsg"
header (see
.Aq Pa netinet/ip_mroute.h )
with the
.Va im_mbz
field set to zero.
Note that this header follows the structure of
.Vt "struct ip"
with the protocol field
.Va ip_p
set to zero.
The IPv6 upcalls have a
.Vt "struct mrt6msg"
header (see
.Aq Pa netinet6/ip6_mroute.h )
with the
.Va im6_mbz
field set to zero.
Note that this header follows the structure of
.Vt "struct ip6_hdr"
with the next header field
.Va ip6_nxt
set to zero.
.Pp
The upcall header contains the
.Va im_msgtype
and
.Va im6_msgtype
fields, with the type of the upcall
.Dv IGMPMSG_*
and
.Dv MRT6MSG_*
for IPv4 and IPv6, respectively.
The values of the rest of the upcall header fields
and the body of the upcall message depend on the particular upcall type.
.Pp
If the upcall message type is
.Dv IGMPMSG_NOCACHE
or
.Dv MRT6MSG_NOCACHE ,
this is an indication that a multicast packet has reached the multicast
router, but the router has no forwarding state for that packet.
Typically, the upcall would be a signal for the multicast routing
user-level process to install the appropriate Multicast Forwarding
Cache (MFC) entry in the kernel.
.Pp
An MFC entry is added by:
.Bd -literal -offset indent
/* IPv4 */
struct mfcctl mc;
memset(&mc, 0, sizeof(mc));
memcpy(&mc.mfcc_origin, &source_addr, sizeof(mc.mfcc_origin));
memcpy(&mc.mfcc_mcastgrp, &group_addr, sizeof(mc.mfcc_mcastgrp));
mc.mfcc_parent = iif_index;
for (i = 0; i \*(Lt maxvifs; i++)
    mc.mfcc_ttls[i] = oifs_ttl[i];
setsockopt(mrouter_s4, IPPROTO_IP, MRT_ADD_MFC,
           (void *)&mc, sizeof(mc));
.Ed
.Bd -literal -offset indent
/* IPv6 */
struct mf6cctl mc;
memset(&mc, 0, sizeof(mc));
memcpy(&mc.mf6cc_origin, &source_addr, sizeof(mc.mf6cc_origin));
memcpy(&mc.mf6cc_mcastgrp, &group_addr, sizeof(mf6cc_mcastgrp));
mc.mf6cc_parent = iif_index;
for (i = 0; i \*(Lt maxvifs; i++)
    if (oifs_ttl[i] \*(Gt 0)
        IF_SET(i, &mc.mf6cc_ifset);
setsockopt(mrouter_s4, IPPROTO_IPV6, MRT6_ADD_MFC,
           (void *)&mc, sizeof(mc));
.Ed
.Pp
The
.Va source_addr
and
.Va group_addr
fields are the source and group address of the multicast packet (as set
in the upcall message).
The
.Va iif_index
is the virtual interface index of the multicast interface the multicast
packets for this specific source and group address should be received on.
The
.Va oifs_ttl[]
array contains the minimum TTL (per interface) a multicast packet
should have to be forwarded on an outgoing interface.
If the TTL value is zero, the corresponding interface is not included
in the set of outgoing interfaces.
Note that for IPv6 only the set of outgoing interfaces can
be specified.
.Pp
An MFC entry is deleted by:
.Bd -literal -offset indent
/* IPv4 */
struct mfcctl mc;
memset(&mc, 0, sizeof(mc));
memcpy(&mc.mfcc_origin, &source_addr, sizeof(mc.mfcc_origin));
memcpy(&mc.mfcc_mcastgrp, &group_addr, sizeof(mc.mfcc_mcastgrp));
setsockopt(mrouter_s4, IPPROTO_IP, MRT_DEL_MFC,
           (void *)&mc, sizeof(mc));
.Ed
.Bd -literal -offset indent
/* IPv6 */
struct mf6cctl mc;
memset(&mc, 0, sizeof(mc));
memcpy(&mc.mf6cc_origin, &source_addr, sizeof(mc.mf6cc_origin));
memcpy(&mc.mf6cc_mcastgrp, &group_addr, sizeof(mf6cc_mcastgrp));
setsockopt(mrouter_s4, IPPROTO_IPV6, MRT6_DEL_MFC,
           (void *)&mc, sizeof(mc));
.Ed
.Pp
The following method can be used to get various statistics per
installed MFC entry in the kernel (e.g., the number of forwarded
packets per source and group address):
.Bd -literal -offset indent
/* IPv4 */
struct sioc_sg_req sgreq;
memset(&sgreq, 0, sizeof(sgreq));
memcpy(&sgreq.src, &source_addr, sizeof(sgreq.src));
memcpy(&sgreq.grp, &group_addr, sizeof(sgreq.grp));
ioctl(mrouter_s4, SIOCGETSGCNT, &sgreq);
.Ed
.Bd -literal -offset indent
/* IPv6 */
struct sioc_sg_req6 sgreq;
memset(&sgreq, 0, sizeof(sgreq));
memcpy(&sgreq.src, &source_addr, sizeof(sgreq.src));
memcpy(&sgreq.grp, &group_addr, sizeof(sgreq.grp));
ioctl(mrouter_s6, SIOCGETSGCNT_IN6, &sgreq);
.Ed
.Pp
The following method can be used to get various statistics per
multicast virtual interface in the kernel (e.g., the number of forwarded
packets per interface):
.Bd -literal -offset indent
/* IPv4 */
struct sioc_vif_req vreq;
memset(&vreq, 0, sizeof(vreq));
vreq.vifi = vif_index;
ioctl(mrouter_s4, SIOCGETVIFCNT, &vreq);
.Ed
.Bd -literal -offset indent
/* IPv6 */
struct sioc_mif_req6 mreq;
memset(&mreq, 0, sizeof(mreq));
mreq.mifi = vif_index;
ioctl(mrouter_s6, SIOCGETMIFCNT_IN6, &mreq);
.Ed
.Ss Advanced Multicast API Programming Guide
Adding new features to the kernel makes it difficult
to preserve backward compatibility (binary and API),
and at the same time to allow user-level processes to take advantage of
the new features (if the kernel supports them).
.Pp
One of the mechanisms that allows preserving the backward
compatibility is a sort of negotiation
between the user-level process and the kernel:
.Bl -enum
.It
The user-level process tries to enable in the kernel the set of new
features (and the corresponding API) it would like to use.
.It
The kernel returns the (sub)set of features it knows about
and is willing to be enabled.
.It
The user-level process uses only that set of features
the kernel has agreed on.
.El
.\"
.Pp
To support backward compatibility, if the user-level process does not
ask for any new features, the kernel defaults to the basic
multicast API (see the
.Sx "Programming Guide"
section).
.\" XXX: edit as appropriate after the advanced multicast API is
.\" supported under IPv6
Currently, the advanced multicast API exists only for IPv4;
in the future there will be IPv6 support as well.
.Pp
Below is a summary of the expandable API solution.
Note that all new options and structures are defined
in
.Aq Pa netinet/ip_mroute.h
and
.Aq Pa netinet6/ip6_mroute.h ,
unless stated otherwise.
.Pp
The user-level process uses new
.Fn getsockopt Ns / Ns Fn setsockopt
options to
perform the API features negotiation with the kernel.
This negotiation must be performed right after the multicast routing
socket is open.
The set of desired/allowed features is stored in a bitset
(currently, in
.Vt uint32_t
i.e., maximum of 32 new features).
The new
.Fn getsockopt Ns / Ns Fn setsockopt
options are
.Dv MRT_API_SUPPORT
and
.Dv MRT_API_CONFIG .
An example:
.Bd -literal -offset 3n
uint32_t v;
getsockopt(sock, IPPROTO_IP, MRT_API_SUPPORT, (void *)&v, sizeof(v));
.Ed
.Pp
This would set
.Va v
to the pre-defined bits that the kernel API supports.
The eight least significant bits in
.Vt uint32_t
are the same as the
eight possible flags
.Dv MRT_MFC_FLAGS_*
that can be used in
.Va mfcc_flags
as part of the new definition of
.Vt "struct mfcctl"
(see below about those flags), which leaves 24 flags for other new features.
The value returned by
.Fn getsockopt MRT_API_SUPPORT
is read-only; in other words,
.Fn setsockopt MRT_API_SUPPORT
would fail.
.Pp
To modify the API, and to set some specific feature in the kernel, then:
.Bd -literal -offset 3n
uint32_t v = MRT_MFC_FLAGS_DISABLE_WRONGVIF;
if (setsockopt(sock, IPPROTO_IP, MRT_API_CONFIG, (void *)&v, sizeof(v))
    != 0) {
    return (ERROR);
}
if (v & MRT_MFC_FLAGS_DISABLE_WRONGVIF)
    return (OK);	/* Success */
else
    return (ERROR);
.Ed
.Pp
In other words, when
.Fn setsockopt MRT_API_CONFIG
is called, the
argument to it specifies the desired set of features to
be enabled in the API and the kernel.
The return value in
.Va v
is the actual (sub)set of features that were enabled in the kernel.
To obtain later the same set of features that were enabled, use:
.Bd -literal -offset indent
getsockopt(sock, IPPROTO_IP, MRT_API_CONFIG, (void *)&v, sizeof(v));
.Ed
.Pp
The set of enabled features is global.
In other words,
.Fn setsockopt MRT_API_CONFIG
should be called right after
.Fn setsockopt MRT_INIT .
.Pp
Currently, the following set of new features is defined:
.Bd -literal
#define	MRT_MFC_FLAGS_DISABLE_WRONGVIF (1 \*(Lt\*(Lt 0)/*disable WRONGVIF signals*/
#define	MRT_MFC_FLAGS_BORDER_VIF   (1 \*(Lt\*(Lt 1)  /* border vif              */
#define MRT_MFC_RP                 (1 \*(Lt\*(Lt 8)  /* enable RP address	*/
#define MRT_MFC_BW_UPCALL          (1 \*(Lt\*(Lt 9)  /* enable bw upcalls	*/
.Ed
.\" .Pp
.\" In the future there might be:
.\" .Bd -literal
.\" #define MRT_MFC_GROUP_SPECIFIC     (1 \*(Lt\*(Lt 10) /* allow (*,G) MFC entries */
.\" .Ed
.\" .Pp
.\" to allow (*,G) MFC entries (i.e., group-specific entries) in the kernel.
.\" For now this is left-out until it is clear whether
.\" (*,G) MFC support is the preferred solution instead of something more generic
.\" solution for example.
.\"
.\" 2. The newly defined struct mfcctl2.
.\"
.Pp
The advanced multicast API uses a newly defined
.Vt "struct mfcctl2"
instead of the traditional
.Vt "struct mfcctl" .
The original
.Vt "struct mfcctl"
is kept as is.
The new
.Vt "struct mfcctl2"
is:
.Bd -literal
/*
 * The new argument structure for MRT_ADD_MFC and MRT_DEL_MFC overlays
 * and extends the old struct mfcctl.
 */
struct mfcctl2 {
        /* the mfcctl fields */
        struct in_addr  mfcc_origin;       /* ip origin of mcasts       */
        struct in_addr  mfcc_mcastgrp;     /* multicast group associated*/
        vifi_t          mfcc_parent;       /* incoming vif              */
        u_char          mfcc_ttls[MAXVIFS];/* forwarding ttls on vifs   */

        /* extension fields */
        uint8_t         mfcc_flags[MAXVIFS];/* the MRT_MFC_FLAGS_* flags*/
        struct in_addr  mfcc_rp;            /* the RP address           */
};
.Ed
.Pp
The new fields are
.Va mfcc_flags[MAXVIFS]
and
.Va mfcc_rp .
Note that for compatibility reasons they are added at the end.
.Pp
The
.Va mfcc_flags[MAXVIFS]
field is used to set various flags per
interface per (S,G) entry.
Currently, the defined flags are:
.Bd -literal
#define	MRT_MFC_FLAGS_DISABLE_WRONGVIF (1 \*(Lt\*(Lt 0)/*disable WRONGVIF signals*/
#define	MRT_MFC_FLAGS_BORDER_VIF       (1 \*(Lt\*(Lt 1) /* border vif          */
.Ed
.Pp
The
.Dv MRT_MFC_FLAGS_DISABLE_WRONGVIF
flag is used to explicitly disable the
.Dv IGMPMSG_WRONGVIF
kernel signal at the (S,G) granularity if a multicast data packet
arrives on the wrong interface.
Usually this signal is used to
complete the shortest-path switch for PIM-SM multicast routing,
or to trigger a PIM assert message.
However, it should not be delivered for interfaces that are not set in
the outgoing interface, and that are not expecting to
become an incoming interface.
Hence, if the
.Dv MRT_MFC_FLAGS_DISABLE_WRONGVIF
flag is set for some of the
interfaces, then a data packet that arrives on that interface for
that MFC entry will NOT trigger a WRONGVIF signal.
If that flag is not set, then a signal is triggered (the default action).
.Pp
The
.Dv MRT_MFC_FLAGS_BORDER_VIF
flag is used to specify whether the Border-bit in PIM
Register messages should be set (when the Register encapsulation
is performed inside the kernel).
If it is set for the special PIM Register kernel virtual interface
(see
.Xr pim 4 ) ,
the Border-bit in the Register messages sent to the RP will be set.
.Pp
The remaining six bits are reserved for future usage.
.Pp
The
.Va mfcc_rp
field is used to specify the RP address (for PIM-SM multicast routing)
for a multicast
group G if we want to perform kernel-level PIM Register encapsulation.
The
.Va mfcc_rp
field is used only if the
.Dv MRT_MFC_RP
advanced API flag/capability has been successfully set by
.Fn setsockopt MRT_API_CONFIG .
.Pp
.\"
.\" 3. Kernel-level PIM Register encapsulation
.\"
If the
.Dv MRT_MFC_RP
flag was successfully set by
.Fn setsockopt MRT_API_CONFIG ,
then the kernel will attempt to perform
the PIM Register encapsulation itself instead of sending the
multicast data packets to user level (inside
.Dv IGMPMSG_WHOLEPKT
upcalls) for user-level encapsulation.
The RP address would be taken from the
.Va mfcc_rp
field
inside the new
.Vt "struct mfcctl2" .
However, even if the
.Dv MRT_MFC_RP
flag was successfully set, if the
.Va mfcc_rp
field was set to
.Dv INADDR_ANY ,
then the
kernel will still deliver an
.Dv IGMPMSG_WHOLEPKT
upcall with the
multicast data packet to the user-level process.
.Pp
In addition, if the multicast data packet is too large to fit within
a single IP packet after the PIM Register encapsulation (e.g., if
its size was on the order of 65500 bytes), the data packet will be
fragmented, and then each of the fragments will be encapsulated
separately.
Note that typically a multicast data packet can be that
large only if it was originated locally from the same hosts that
performs the encapsulation; otherwise the transmission of the
multicast data packet over Ethernet for example would have
fragmented it into much smaller pieces.
.\"
.\" Note that if this code is ported to IPv6, we may need the kernel to
.\" perform MTU discovery to the RP, and keep those discoveries inside
.\" the kernel so the encapsulating router may send back ICMP
.\" Fragmentation Required if the size of the multicast data packet is
.\" too large (see "Encapsulating data packets in the Register Tunnel"
.\" in Section 4.4.1 in the PIM-SM spec
.\" draft-ietf-pim-sm-v2-new-05.{txt,ps}).
.\" For IPv4 we may be able to get away without it, but for IPv6 we need
.\" that.
.\"
.\" 4. Mechanism for "multicast bandwidth monitoring and upcalls".
.\"
.Pp
Typically, a multicast routing user-level process would need to know the
forwarding bandwidth for some data flow.
For example, the multicast routing process may want to time out idle MFC
entries, or for PIM-SM it can initiate (S,G) shortest-path switch if
the bandwidth rate is above a threshold for example.
.Pp
The original solution for measuring the bandwidth of a dataflow was
that a user-level process would periodically
query the kernel about the number of forwarded packets/bytes per
(S,G), and then based on those numbers it would estimate whether a source
has been idle, or whether the source's transmission bandwidth is above a
threshold.
That solution is far from being scalable, hence the need for a new
mechanism for bandwidth monitoring.
.Pp
Below is a description of the bandwidth monitoring mechanism.
.Bl -bullet
.It
If the bandwidth of a data flow satisfies some pre-defined filter,
the kernel delivers an upcall on the multicast routing socket
to the multicast routing process that has installed that filter.
.It
The bandwidth-upcall filters are installed per (S,G).
There can be
more than one filter per (S,G).
.It
Instead of supporting all possible comparison operations
(i.e., \*(Lt \*(Lt= == != \*(Gt \*(Gt= ), there is support only for the
\*(Lt= and \*(Gt= operations,
because this makes the kernel-level implementation simpler,
and because practically we need only those two.
Furthermore, the missing operations can be simulated by secondary
user-level filtering of those \*(Lt= and \*(Gt= filters.
For example, to simulate !=, then we need to install filter
.Dq bw \*(Lt= 0xffffffff ,
and after an
upcall is received, we need to check whether
.Dq measured_bw != expected_bw .
.It
The bandwidth-upcall mechanism is enabled by
.Fn setsockopt MRT_API_CONFIG
for the
.Dv MRT_MFC_BW_UPCALL
flag.
.It
The bandwidth-upcall filters are added/deleted by the new
.Fn setsockopt MRT_ADD_BW_UPCALL
and
.Fn setsockopt MRT_DEL_BW_UPCALL
respectively (with the appropriate
.Vt "struct bw_upcall"
argument of course).
.El
.Pp
From an application point of view, a developer needs to know about
the following:
.Bd -literal
/*
 * Structure for installing or delivering an upcall if the
 * measured bandwidth is above or below a threshold.
 *
 * User programs (e.g. daemons) may have a need to know when the
 * bandwidth used by some data flow is above or below some threshold.
 * This interface allows the userland to specify the threshold (in
 * bytes and/or packets) and the measurement interval. Flows are
 * all packet with the same source and destination IP address.
 * At the moment the code is only used for multicast destinations
 * but there is nothing that prevents its use for unicast.
 *
 * The measurement interval cannot be shorter than some Tmin (3s).
 * The threshold is set in packets and/or bytes per_interval.
 *
 * Measurement works as follows:
 *
 * For \*(Gt= measurements:
 * The first packet marks the start of a measurement interval.
 * During an interval we count packets and bytes, and when we
 * pass the threshold we deliver an upcall and we are done.
 * The first packet after the end of the interval resets the
 * count and restarts the measurement.
 *
 * For \*(Lt= measurement:
 * We start a timer to fire at the end of the interval, and
 * then for each incoming packet we count packets and bytes.
 * When the timer fires, we compare the value with the threshold,
 * schedule an upcall if we are below, and restart the measurement
 * (reschedule timer and zero counters).
 */

struct bw_data {
        struct timeval  b_time;
        uint64_t        b_packets;
        uint64_t        b_bytes;
};

struct bw_upcall {
        struct in_addr  bu_src;         /* source address            */
        struct in_addr  bu_dst;         /* destination address       */
        uint32_t        bu_flags;       /* misc flags (see below)    */
#define BW_UPCALL_UNIT_PACKETS (1 \*(Lt\*(Lt 0) /* threshold (in packets)    */
#define BW_UPCALL_UNIT_BYTES   (1 \*(Lt\*(Lt 1) /* threshold (in bytes)      */
#define BW_UPCALL_GEQ          (1 \*(Lt\*(Lt 2) /* upcall if bw \*(Gt= threshold */
#define BW_UPCALL_LEQ          (1 \*(Lt\*(Lt 3) /* upcall if bw \*(Lt= threshold */
#define BW_UPCALL_DELETE_ALL   (1 \*(Lt\*(Lt 4) /* delete all upcalls for s,d*/
        struct bw_data  bu_threshold;   /* the bw threshold          */
        struct bw_data  bu_measured;    /* the measured bw           */
};

/* max. number of upcalls to deliver together */
#define BW_UPCALLS_MAX				128
/* min. threshold time interval for bandwidth measurement */
#define BW_UPCALL_THRESHOLD_INTERVAL_MIN_SEC	3
#define BW_UPCALL_THRESHOLD_INTERVAL_MIN_USEC	0
.Ed
.Pp
The
.Vt bw_upcall
structure is used as an argument to
.Fn setsockopt MRT_ADD_BW_UPCALL
and
.Fn setsockopt MRT_DEL_BW_UPCALL .
Each
.Fn setsockopt MRT_ADD_BW_UPCALL
installs a filter in the kernel
for the source and destination address in the
.Vt bw_upcall
argument,
and that filter will trigger an upcall according to the following
pseudo-algorithm:
.Bd -literal
 if (bw_upcall_oper IS "\*(Gt=") {
    if (((bw_upcall_unit & PACKETS == PACKETS) &&
         (measured_packets \*(Gt= threshold_packets)) ||
        ((bw_upcall_unit & BYTES == BYTES) &&
         (measured_bytes \*(Gt= threshold_bytes)))
       SEND_UPCALL("measured bandwidth is \*(Gt= threshold");
  }
  if (bw_upcall_oper IS "\*(Lt=" && measured_interval \*(Gt= threshold_interval) {
    if (((bw_upcall_unit & PACKETS == PACKETS) &&
         (measured_packets \*(Lt= threshold_packets)) ||
        ((bw_upcall_unit & BYTES == BYTES) &&
         (measured_bytes \*(Lt= threshold_bytes)))
       SEND_UPCALL("measured bandwidth is \*(Lt= threshold");
  }
.Ed
.Pp
In the same
.Vt bw_upcall ,
the unit can be specified in both BYTES and PACKETS.
However, the GEQ and LEQ flags are mutually exclusive.
.Pp
Basically, an upcall is delivered if the measured bandwidth is \*(Gt= or
\*(Lt= the threshold bandwidth (within the specified measurement
interval).
For practical reasons, the smallest value for the measurement
interval is 3 seconds.
If smaller values are allowed, then the bandwidth
estimation may be less accurate, or the potentially very high frequency
of the generated upcalls may introduce too much overhead.
For the \*(Gt= operation, the answer may be known before the end of
.Va threshold_interval ,
therefore the upcall may be delivered earlier.
For the \*(Lt= operation however, we must wait
until the threshold interval has expired to know the answer.
.Sh EXAMPLES
.Bd -literal -offset indent
struct bw_upcall bw_upcall;
/* Assign all bw_upcall fields as appropriate */
memset(&bw_upcall, 0, sizeof(bw_upcall));
memcpy(&bw_upcall.bu_src, &source, sizeof(bw_upcall.bu_src));
memcpy(&bw_upcall.bu_dst, &group, sizeof(bw_upcall.bu_dst));
bw_upcall.bu_threshold.b_data = threshold_interval;
bw_upcall.bu_threshold.b_packets = threshold_packets;
bw_upcall.bu_threshold.b_bytes = threshold_bytes;
if (is_threshold_in_packets)
    bw_upcall.bu_flags |= BW_UPCALL_UNIT_PACKETS;
if (is_threshold_in_bytes)
    bw_upcall.bu_flags |= BW_UPCALL_UNIT_BYTES;
do {
    if (is_geq_upcall) {
        bw_upcall.bu_flags |= BW_UPCALL_GEQ;
        break;
    }
    if (is_leq_upcall) {
        bw_upcall.bu_flags |= BW_UPCALL_LEQ;
        break;
    }
    return (ERROR);
} while (0);
setsockopt(mrouter_s4, IPPROTO_IP, MRT_ADD_BW_UPCALL,
          (void *)&bw_upcall, sizeof(bw_upcall));
.Ed
.Pp
To delete a single filter, use
.Dv MRT_DEL_BW_UPCALL ,
and the fields of bw_upcall must be set to
exactly same as when
.Dv MRT_ADD_BW_UPCALL
was called.
.Pp
To delete all bandwidth filters for a given (S,G), then
only the
.Va bu_src
and
.Va bu_dst
fields in
.Vt "struct bw_upcall"
need to be set, and then just set only the
.Dv BW_UPCALL_DELETE_ALL
flag inside field
.Va bw_upcall.bu_flags .
.Pp
The bandwidth upcalls are received by aggregating them in the new upcall
message:
.Bd -literal -offset indent
#define IGMPMSG_BW_UPCALL  4  /* BW monitoring upcall */
.Ed
.Pp
This message is an array of
.Vt "struct bw_upcall"
elements (up to
.Dv BW_UPCALLS_MAX
= 128).
The upcalls are
delivered when there are 128 pending upcalls, or when 1 second has
expired since the previous upcall (whichever comes first).
In an
.Vt "struct upcall"
element, the
.Va bu_measured
field is filled in to
indicate the particular measured values.
However, because of the way
the particular intervals are measured, the user should be careful how
.Va bu_measured.b_time
is used.
For example, if the
filter is installed to trigger an upcall if the number of packets
is \*(Gt= 1, then
.Va bu_measured
may have a value of zero in the upcalls after the
first one, because the measured interval for \*(Gt= filters is
.Dq clocked
by the forwarded packets.
Hence, this upcall mechanism should not be used for measuring
the exact value of the bandwidth of the forwarded data.
To measure the exact bandwidth, the user would need to
get the forwarded packets statistics with the
.Fn ioctl SIOCGETSGCNT
mechanism
(see the
.Sx Programming Guide
section) .
.Pp
Note that the upcalls for a filter are delivered until the specific
filter is deleted, but no more frequently than once per
.Va bu_threshold.b_time .
For example, if the filter is specified to
deliver a signal if bw \*(Gt= 1 packet, the first packet will trigger a
signal, but the next upcall will be triggered no earlier than
.Va bu_threshold.b_time
after the previous upcall.
.\"
.Sh SEE ALSO
.Xr getsockopt 2 ,
.Xr recvfrom 2 ,
.Xr recvmsg 2 ,
.Xr setsockopt 2 ,
.Xr socket 2 ,
.Xr icmp6 4 ,
.Xr inet 4 ,
.Xr inet6 4 ,
.Xr intro 4 ,
.Xr ip 4 ,
.Xr ip6 4 ,
.Xr pim 4 ,
.Xr mrouted 8 ,
.Xr sysctl 8
.\"
.Sh AUTHORS
.An -nosplit
The original multicast code was written by
.An David Waitzman
(BBN Labs),
and later modified by the following individuals:
.An Steve Deering
(Stanford),
.An Mark J. Steiglitz
(Stanford),
.An Van Jacobson
(LBL),
.An Ajit Thyagarajan
(PARC),
.An Bill Fenner
(PARC).
.Pp
The IPv6 multicast support was implemented by the KAME project
.Pq Lk http://www.kame.net ,
and was based on the IPv4 multicast code.
The advanced multicast API and the multicast bandwidth
monitoring were implemented by
.An Pavlin Radoslavov
(ICSI)
in collaboration with
.An Chris Brown
(NextHop).
.Pp
This manual page was written by
.An Pavlin Radoslavov
(ICSI).