Multicast Solution – Redundant BSR and RP

As we discussed many things about Multicast in previous posts, it will be better to see some real time solution which will help us for better understanding. How multicast is deployed in large enterprises.

Test Requirements – Large Enterprise is having 2 redundant multicast servers with same multicast group 239.1.1.1. As any of multicast servers went faulty due to any reason, other multicast server will work as backup to serve network and avoid downtime. Same time they want redundant BSR and RP for network. In any case Primary is not available, that time backup should take care.

Test Topology

Testing description:

Customer requires redundancy for Multicast servers and RP. Master multicast server goes down due to any reason, and then backup server sends the same multicast traffic 239.1.1.1. We need to check the convergence time how long time the client side can receive the 239.1.1.1 and same time If Primary RP goes down, Backup RP should work as primary and spoke location can access backup multicast server.

Test Process:

Multicast host 6.1.1.2 wants to join multicast group 239.1.1.1. MP 2816 router check the RP mapping and send the request to Primary RP. As source receive Join request from host, Now 4.1.1.2 (Multicast source) sends the multicast traffic 239.1.1.1 to Multicast host 6.1.1.2, It  can receive the multicast stream properly.

Now 4.1.1.2 (Multicast source) is down due to some issue.

OR

Primary RP is down due to any reason. That time

Backup RP will work as Master and backup multicast server 5.1.1.2 send the same 239.1.1.1 multicast stream manually.

And finally the convergence time less than 5 second, Now Multicast host can receive same multicast stream 239.1.1.1.This solution is majorly deployed to achieve redundancy in multicast network.  

Test Configurations:

MP1800-36:

ip mef

ip load-sharing per-destination

ip multicast-routing

controller e1 1/0

 channel-group 0 timeslots 1-31

 framing crc4

 exit

!end

vlan 1

 exit

interface fastethernet0

 ip address 2.1.1.1 255.255.255.0

 ip pim sparse-mode

 exit

interface fastethernet1

 ip address 4.1.1.1 255.255.255.0

 ip pim sparse-mode

 exit

interface serial1/0:0

 encapsulation hdlc

 ip address 1.1.1.2 255.255.255.0

 bandwidth 1984

 ip pim sparse-mode

 exit

                               

router ospf 1

 network 0.0.0.0 255.255.255.255 area 0

 exit

ip pim bsr-candidate fastethernet0 10    ### Primary BSR ###

ip pim rp-candidate fastethernet0 190   ### Primary RP ###

MP801E:

ip multicast-routing

controller e1 0/0

 channel-group 0 timeslots 1-31

 framing crc4

 exit

interface fastethernet0

 ip address 3.1.1.1 255.255.255.0

 ip pim sparse-mode

 exit

interface fastethernet1

 ip address 5.1.1.1 255.255.255.0

 ip pim sparse-mode

 exit

interface serial0/0:0

 encapsulation hdlc

 ip address 1.1.1.1 255.255.255.0

 bandwidth 1984

 ip pim sparse-mode

 exit

                                 

router ospf 1

 network 0.0.0.0 255.255.255.255 area 0

 exit

ip pim bsr-candidate fastethernet0

ip pim rp-candidate fastethernet0

MP2816:

ip mef

ip load-sharing per-destination

ip multicast-routing

vlan 1

 description default

 port 0-23 untagged

 exit

interface fastethernet0

 ip address 3.1.1.2 255.255.255.0

 ip pim sparse-mode

 exit

interface fastethernet1

 ip address 2.1.1.2 255.255.255.0

 ip pim sparse-mode

 exit

            

interface switchethernet1

 ip address 6.1.1.1 255.255.255.0

 vlan 1

 ip pim sparse-mode

 exit

                            

router ospf 1

 network 0.0.0.0 255.255.255.255 area 0

 exit

### MP2816 router is spoke router it will learn BSR information with PIM messages and RP information from BSR. ###

Test Results:

1. As Primary Multicast server goes down and Primary Master RP is still live, That time backup multicast server will register with Primary RP. It will send same multicast traffic 239.1.1.1 in network.
2. As Primary multicast server 239.1.1.1 and Primary RP (MP1800) goes down. That time Backup RP (MP 801E) will work as Primary RP and backup Multicast server will register to available RP and it will send same multicast stream(239.1.1.1) in network.

That means this multicast solution will overcome both challenges, which normally enterprises has to deal.

Hope you like this solution. This above configuration is tested on Maipu Routers, Cisco routers will have same configurations.

Your comments will help me to improve my posts. For any queries put comments with your mail id…

Multicast - Bootstrap Router (BSR)

As we discussed in last post about BSR in PIM sparse mode, Let’s see more detail about BSR in this section...

BSR is standard protocol, which work with PIM to advertise RP in network. But we need to elect BSR first. We can define BSR as -BSR is Router, Which will collect all RP advertisement from network and make a set of RP address and advertise/share in network.  In Cisco Auto-RP - it is different, Auto-RP will select the best RP and advertise only one best RP in all network.

                Both protocols use the same information which is coming from rp candidate advertisement.

Let me explain this process- First you will select a router. Which you want to designate as BSR

Syntax – ip pim bsr- candidate fastethernet0

Using this command we can designate the router as BSR router in network. Then those routers which are near to source or concern router we want to designate as RP. We need to configure this command.

Syntax- ip pim rp-candidate fastethernet0

At this time, router will start sending candidate RP announcements to BSR Routers. BSR will collect the announcements and make a set of all RP announcements and advertise in network through PIM messages.

                BSR advertisement will go thru RPF check same like PIM. Then all routers will have BSR information in network along with RP information. As the join request will come to router, Router will check the RP address and reach to source. BSR is reducing manual overhead to configure static RP in every router.

                          

             We can configure redundant BSR in network. In that case we need to configure priority for BSR. Highest priority will win and elected as Master BSR. As you enable a router as BSR, it will send bootstrap messages in network.BSR message contains BSR address with priority. As other router will receive the bootstrap messages it will come to know that one more BSR router is in network. It will check the priority associated with message and select our role. Depending on priority master will be elected. Higher wins...

Syntax – ip pim bsr-candidate interface [hash-mask-length [priority]]

Priority – 0 to 255

In Maipu Routers - By default, hash-mask-length is 10 and priority is 0.

In Cisco Routers – by default the priority is 0.

                                    In RP also we can configure priority, In RP lowest priority wins. As we require RP for particular group that time we can configure Access-list and bind it with RP and advertise with priority. In case of lower priority BSR will consider that RP and advertise in network.

Default RP priority in Maipu Router – 192

Default RP priority in Cisco - 0 which specifies 192.

                   

                          If some RP is advertising lowest priority like 190, then it will master RP which BSR will be advertise in network.

                          RP send unicast messages to configured BSR routers continuously. So as the RP change happen in network. It will get notified by BSR immediately.

Hope this post clarified you about BSR and RP election and working flow. We will see configuration and scenario for more understanding in coming posts...

If you like this post and for any queries Plz comment with your mail id...