Manual:Routing/Multicast

From CableFree RadioOS
Jump to navigation Jump to search
Version.png

Applies to RadioOS: v3.x, v4.x

Summary

Protocol Independent Multicast - Sparse Mode (PIM-SM or PIM) enables RadioOS to support multicast streaming over network area where routers have PIM set up. Several configured PIM routers together will make multicast cloud where client devices can use IGMP to manage subscriptions to streams. PIM should be used when network topology is complex or stream sources are connected to multicast cloud. Continuous cloud must have configured unique rendezvous point for multicast group or groups used in it and other participants should know how to reach rendezvous point. In simple case when in part of cloud reside only potential clients and no stream sources IGMP proxy can be used instead to conserve resources.

Requirements

Multicast is available on all architectures supported by RadioOS. Packages required:

  • system
  • multicast
Icon-note.png

Note: v3.x routing-test and multicast packages are incompatible. In case when both are present one of them will be disabled


Icon-note.png

Note: To get the package you have to download all-packages archive and upload/install multicast package separately on the router


Protocol independent multicast (PIM)

Menu: /routing pim

General PIM protcol settings.

Property Desciption
switch-to-spt (yes|no, default: yes) whether to switch to Shortest Path Tree (SPT) phase if multicast data bandwidth threshold is reached. For routers upstream from RP, if this option is disabled, it means that the router will not proceed from protocol phase one (register encapsulation) to native multicast traffic flow. It is recommended to enable this option.
switch-to-spt-bytes (integer, default: 0) multicast data bandwidth threshold. If this threshold is reached in the specified time interval, switching to Shortest Path Tree (SPT) happens. If value 0 is configured, switching will happen immediately.
switch-to-spt-interval (time, default: 100s) time interval in which to account multicast data bandwidth, used in conjunction with switch-to-spt-bytes to determine if switching threshold is reached.

Interfaces

Menu: /routing pim interface

Since RadioOS v4.6 it is possible to specify source address interface will use to participate in multicast cloud. Previously one of interface addresses was chosen without any particular order.

Configuration of interface of the router that will participate in multicast network. Interfaces that are not configured here (or in IGMP-Proxy) will discard multicast packets.

When deploying multicast configuration over wireless links one should be cautious how and what works. For details about multicast and wireless links.

Icon-note.png

Note: There is no interface count limitation in this menu other than how much hardware can handle



Property Desciption
alternative-subnets (IP address/mask Default:  nil) : if router can receive multicast streams over groups that are not in standard Class-D section then you have to set up this field, so these groups are recognised as multicast groups and will not be discarded.
assert-override-interval (time Default: 3s) time that is subtracted by assert winner from assert-time field, to ensure, that assert winner will always send its assert messages before everyone else.
assert-time (time Default: 3m) Time interval when assert-winner will send out repeated assert.
comment (text) text information for the entry
copy-from (number) use other, already configured entry as stencil for this new one
disabled (yes | no; Default: no) state of the entry
dr-priority (integer Default: 1) if for stream source more than one router with multicast support is available, then one with highest priority will become Designated router of that multicast stream and will handle stream delivery to RP. Higher value means higher priority.
hello-holdtime (time Default: 1m45s) how long consider sender of hello packet received on interface in neighbour list. (usually 3.5 times of hello-period)
hello-period (time Default: 30s) how often hello packet will be sent over this interface.
hello-trigerred-delay (time Default: 5s) when interface starts to participate in multicast cloud then this value is max time interface will wait before sending hello packet. That period of waiting is random from 0 to value set in this field.
igmp-version (IGMPv1 | IGMPv2 | IGMPv3 Default: IGMPv2) what IGMP protocol version to support on the interface.
interface (interface name) interface name that will participate in multicast cloud with these settings.
join-prune-holdtime (time Default: 3m30s) how long save join or prune status before discard it
join-prune-period (time Default: 1m) time interval between sending out join or prune messages
preferred-source-address (IP address Default: 0.0.0.0) address that should be used to send out IGMP/PIM packets. Address used should be already assigned to interface. (introduced in 4.6)
propagation-delay (integer in milliseconds Default: 50) expected propagation delay between PIM routers on this etwork or link.
protocols (pim, igmp; Default: 'pim,igmp) what protocols to support on the interface
require-hello (yes|no Default:yes) if sending PIM router have to be neighbour to receiving (this) router to work with those packets.
tracking-support (yes | no Default: yes) if propagation-delay is not negotiated or is not set then that value will be suppressed, if one of PIM neighbours has set false in this field, then propagation-delay will be suppressed.
override-interval (integer in milliseconds Default: 250) will override propagation-delay negotiated value if set delay time is smaller than this.

Rendezvous point

Menu: /routing pim rp

Rendezvous point configuration. Rendezvous point (RP) is a distribution point for multicast group, source provides its data to it, and if there are any subscribers, then RP will provide data to client. Note, that RP will always receive data stream if that exists.

Property Desciption
comment (text) add comment to static RP entry
copy-from (number) creates another RP just like one you pointed to with number you used.
disabled (yes, no) used to change status of RP entry effectively disabling or enabling it.
group (multicas group address Default: 224.0.0.0/4) sets what group this RP will be assigned to. Values accepted are class D ip addresses with mask, thus effectively marking multiple groups to this RP entry e.g. 224.10.10.0/24 will add 256 groups starting with 224.10.10.0 till 224.10.10.255.
hash-mask-length (number 4..32 Default: 30) when multicast group have multiple RPs, and they are same scope and same priority, then this value is compared. and so you can load balance this way.
priority (number Default: 192) if several RPs are available for multicast group, and they are both with same scope, then RP with highest priority is chosen. Smaller non-negative value is considered of higher priority. Example: priority of 100 is higher than priority of 101.
address (IP address) at what address you have to look for RP for multicast group specified in group field. If group is set to one of routers interfaces, it should be reachable through whole multicast network, if it not, you will have to set up rules in MRIB (multicast routing information base).

Rendezvous point candidates

Menu: /routing pim rp-candidates

Rendezvous point candidate configuration, when RP is elected for multicast group

Property Desciption
comment (text) additional textual information about entry
copy-from (number) create this entry using other entry as a stencil
disabled (yes | no Default: no) state of entry
group (multicast group address Default: 224.0.0.0/4) routes with will be chosen to be a group RP if no other RP will not participate with higher priority
holdtime (time Default: 2m30s) after what time next election will be initiated
is-scope-zone (yes|no Default: no) if set to yes, scope-zone setting is obeyed, if set to no, then scope-zone just represents range of groups that it will function as RP
priority (number Default: 192) value is used when RP is elected, lower value mean higher priority
interface (interface) to what interface to bind to if this router is elected as multicast groups RP

Bootstrap router candidates

Menu: /routing pim bsr-candidates

bootstrap router candidate configuration

Property Desciption
comment (text) set text describing bsr-candidate list entry
disabled (yes|no Default: no) set state of the list entry
hash-mask-length (number 4..32 Default: 30) to how much first bits of multicast group should be hashed to reduce protocol overhead
is-scope-zone (yes|no Default: no) if set to yes, scope-zone setting is obeyed, if set to no, then scope-zone just represents range of groups that it will function as BSR
priority (number Default: 1) priority of the router in bsr election
scope-zone (IP address/mask Default: 224.0.0.0/4) multicast group range that this router will function as BSR
interface (interface) interface of the router that bsr-candidate will be attached to and if elected BSR

Multicast route information base

Menu: /routing pim mrib

MRIB routes are used for reverse path forwarding check. In a way, they perform opposite function that FIB (Forwarding Information Base) routes: FIB is used to find the right By default, MRIB is populated by FIB routes. Use "multicast" routing filter chanin to control that or set specific parameters for imported FIB routes (e.g. you can change the distance of the route). In addition, you can specify static MRIB routes. This is useful only if you are using multihoming and multicast packet flow will be different from unicast packet flow.

Active MRIB entries that are imported from FIB are shown with "dynamic" flag.

Property Desciption
comment (text) textual note to entry can be added to static entries only.
copy-from (number) use other entry as a template to create new one.
destination (IP address/mask Default: 0.0.0.0/0) hosts that will be reachable through gateway
disabled (yes|no Default: no) status of entry
metric (integer Default: 1) value of cost of the route. Route with least weight will be used if available.
gateway (IP address) address through where hosts listed in destination field will be reachable.

IGMP group status

Menu: /routing pim igmp-group

Property Desciption
interface (interface) group join request received/sent on this interface
group (IP address) IGMP group of interest
source (IP address) source of IGMP request
state (exclude | forward | don't forward) state of IGMP group membership
version (IGMPv1|IGMPv2|IGMPv3) version of IGMP protocol used
timeout (time) time when entry will expire if not refreshed

Multicast neighbors

Menu: /routing pim neighbors

This menu only allows to see information about multicast routers that are reachable within one Ethernet from all interfaces participating in multicast routing. This list is created and updated automatically according to state of multicast network.

Property Desciption
address (ip address) IP address of neighbour multicast router that router have received hello packet.
interface (text) on what interface hello packet was received
priority (number 1..255) priority of the neighbour router
holdtime (time) how long entry will be held in neighbour list (configured in interface menu hello-holdtime)
timeout (time) how much time left when entry will be dropped from list if no hello packets are received. Every time hello packet is received this entry will be refreshed.

Bootstrap router status

Menu: /routing pim bsr

Property Desciption
zone-type (active | expiring | configured ) type of the zone
valign="top"|bsr-address (IP address) address of BSR router
scope-zone (IP address/mask) multicast group range this router is a BSR
bsr-priority (integer) priority of BSR router
local-address (IP address) local BSR candidate address in scope zone
local-priority (integer) local BSR candidate priority in scope zone
state (init | candidate | pending | elected | no-info | accept-any | accept-preferred) state of BSR router
timeout (time | -1) time-out when this entry will be removed
  • -1 : never expire
  • time value : time remaining to expiry
sz-timeout (time | -1) in what time when sope zone will time out.
  • -1 : never expire
  • time value : time remaining to expiry

Multicast forwarding cache status

Menu: /routing pim mfc

Multicast forwarding cache - this section only provides information about current state of multicast cloud at given router, showing states of joins for multicast groups.

Property Desciption
group (IP address) name of multicast group
source (IP address) source of multicast group data
rp (IP address) address of RP for that multicast group
incoming-interface (interface) interface that is used to receive multicast data stream
outgoing-interface (interface) interface that is used to transmit multicast data stream

Multicast group joins status

Menu: /routing pim join

Group join list of all the group joins that are registered by PIM-SM Multicast forwarding cache - this section only provides information about current state of multicast cloud at given router, showing states of joins for multicast groups.

Property Desciption
group (IP address) multicast group that has at least one registered join request
source (IP address) data provider for that group
rp (IP address) rendezvous point for multicast group
upstream-interface-source (interface) router interface receives data stream of the multicast group
upstream-interface-rp (interface) router interface that is toward the rendezvous point
upstream-mrib-nexthop (IP address) address of the next router towards RP
upstream-pim-nexthop (IP address) address of next router towards RP according to PIM RP tree
join-state (joined | not-joined | rpt-not-joined | rpt-pruned | rpt-not-pruned) state of this entry towards RP
join-register-state (joined | pruned | join-pending | unknown) join status on register interface
timeout (time) time-out when entry will be removed from the list.
keepalive-timer (yes|no) how long entry will be kept in the list
i-am-designated-router (interface list) interface name list on which router is chosen as designated router
local-receivers (interface list) interfaces where are clients registered with (*.G) join
joined-rp (interface list) list of interfaces that have clients that originated (*,*,RP) join
joined-wc (interface list) list of interfaces that have clients that originated (*.G) join
joined (interface list) list of interfaces that are in joined state
pruned (interface list) list of interfaces that are in prune state
prune-pending () list of interfaces that are in prune-pending state
assert-winner (interface list) list of interfaces that are in assert-winner state
assert-looser (interface list) list of interfaces that are assert-lost state
assert-winner-wc (interface list) list of interfaces that have (*,G) join and have assert-winner state
assert-looser-wc (interface list) list of interfaces that have (*,G) join and have assert-lost state
assert-tracking-wc (interface list) list of interfaces that have (*,G) join and will track assert
could-assert-wc (interface list) list of interfaces that have (*,G) join and could trigger assert
immediate-rp (interface list) list of interfaces that are included in the immediate outgoing interfaces for the corresponding (*,*,RP) entry.
immediate-wc (interface list) list of interfaces that are included in the immediate outgoing interfaces for the corresponding (*,RP) entry.
immediate-sg (interface list) list of interfaces that are included in the immediate outgoing interfaces for the corresponding (S,G) entry.
immediate-sg-rpt (interface list) list of interfaces that are included in the immediate outgoing interfaces for the corresponding (S,G,rpt) entry.
include-wc (interface list) list of interfaces to which traffic might be forwarded because of hosts that are local members on that interface.

See also