Thursday, November 12, 2009
OSPF Broadcast CCNA Lab
Today I spent a few hours playing with an OSPF broadcast lab which is basically an inter-network within a LAN instead of the usual point-to-point WAN. The main purpose of this lab was for me to determine which router would become the DR (Designated Router) that would be responsible for the LSA's. I created the network using the 192.168.146.0 /26 network and set the RID's up in the way where R1 would be the DR, R4 would be the BDR (Backup Designated Router), and the other two would be DRother's (non-DR's). I spent some time confused because I would notice that even if R1 had the highest R.I.D. (18.104.22.168) it still wouldn't become the DR for this network. After I did a little research it hit me that it all depends on which routers comes up first because with OSPF when a new router comes up within the OSPF network, the network won't change it's DR settings regardless if the new router has a higher RID or not. The best way to get around this without having to reload the routers in a tedious particular order is to run the clear ip ospf process command. This basically restarts OSPF on the network and will force all the routers to form fully adjacent connections with each other b.ased on the routers currently running OSPF within that area.
I was forced to think about how OSPF functions using LSA's in a single area which was good because this is something you could run into on a production network. I didn't do a multi area network lab today because it really isn't much to it CCNA wise. You do have to understand the area 0 is the backbone area while other area are seperated with an ABR (Area Border Router). An ABR is exactly how you would think it would be it sits on the border between different OSPF areas to allow the different areas to communicate while preventing the areas from having to share LSDB info which would cause more processing power and slower convergence due to the area being larger. Breaking large OSPF networks into smaller areas is always a good idea, especially when there are 100's if not 1000's or routers and subnets in the Autonomous System. Different areas allows the network to be more scalable and not have to converge every time a single router interface is changed!