R&S Quick Notes – Multicast & IPv6

Multicast

  • BSR is also commonly referred to as PIMv2.
  • Pay special attention to when using Frame-Relay non-broadcast types. Multicast will not work. Tunnels might be needed.
  • BSR – when multiple c-RP announces same groups, a longer match will be used to determine the RP, regardless of the RP priority set.
  • With TTL scoping, if the Packet TTL >= Interface TTL, then the packet is forwarded, else dropped.
  • GRE-tunnel –  If the unicast source is reachable via tunnel, a RPF failure will occur. Correct with Mroute.
  • Know how to spot RPF failures.
  • Multicast Filtering:

1. Q – Prevent PIM neighbor establishments, but allow IGMP  client joins?

A – On Central router : “ip pim neighbors filter” & the Stub router : “ip igmp helper-address”

2. Q – Filter specific multicast groups, while still allowing IGMP traffic?

A – “ip multicast boundary {acl}”

3. Q – Deny clients from joining specific multicast groups?

A – “ip igmp access-group {acl}”

4. Q – Statically filter RP requests and responses, (no Auto-RP, no BSR)?

A – “ip pim rp address {IP} {acl}”

5. Q – Client RP filtering, Limit join/prune messages for specific RP’s?

A – “ip pim accept-rp {RP-IP/auto-RP} {acl}”

6. Q – Auto-RP – Limit the Candidate-RP’s announcements?

A – “ip pim send-rp-announce {int} scope {no} group-list {acl}”

7. Q – Auto-RP – Limit what mgroups a MA accept from specific RP’s?

A – “ip pim rp-announce-filter rp-list {acl} group-list {acl}”

8. Q – Filter the BSR messages on a interface?

A – “ip pim bsr-border”

9. Q – Limit the amount of multicast routes in the mrouting table?

A – “ip multicast route-limit”

10. Q – Limit the rate a source can sent traffic to a group on a interface?

A – “ip multicast rate-limit group-list {acl} {kbps}”

IPv6

  • RIPng – “no ip split-horizon” in a process command not a interface command.
  • EIGRPv6 – Do not forget to enable eigrp under the process.
  • IPv6 tunnel method with least overhead : IPv6IP
  • Tunnel protocol numbers for ACL’s : IPv6IP = Protocol-41,  &  GRE IPv6 = Protocol-47
  • You can not redistribute a default static route(::/0) with ospfv3.
  • Dynamic information (ie IGP next-hops)  recurses to remote link-local address, not the global unicast interface.
Advertisement

Converting IPv4 to IPv6 and back

Converting from IPv4 to IPv6

is so easy, yet everyone seem to convert a IPv4 address to binary, then to IPv6. Why? Why waste time and do things the long way? Not cool.

When would you need to do this? One specific use is IPv6 6-to-4 tunnels, which always concatenates 2002::/16 with the IPv4 address embedded.
With Automatic 6-to-4-tunnels, your address format is as follow:
2002:<32 bit IPv4 site address in Hex>:<16 bit network number in Hex>::/64

The question is how to do the conversion.

Firstly before starting I will assume everyone knows the following:

  • Binary is a Base-2 numbering system, as it has only 0,1
  • Decimal is a Base-10 numbering system, as it has 0,1,2,3,4,5,6,7,8,9
  • Hexadecimal is a Base-16 numbering system, as it has 0,1,2,3,4,5,6,7,8,9,A,B,C,D,E,F

I also assume you know the hex values in decimal:

A = 10
B = 11
C = 12
D = 13
E = 14
F = 15

Two more things I would like to mention before explaining the conversion.
An IPv4 address : example 192.168.99.1

  • Each Octet (8 bits) “between the dot-thingys” denote 1 byte

An IPv6 address : example 2001:0db8:85a3:0000:0000:8a2e:0370:7334

  • Two Tuples (1 Tuple = 4 bits = 1 Hex character) denotes 1 byte

Then converting is easy. Lets take the following IPv4 address : 192.168.99.1 and convert it to Hex.

Continue reading “Converting IPv4 to IPv6 and back”