You're creating a multi-AS BGP test lab on Sunday evening;
The core AS is running 12.2SRC code;
You insert a P-router in the core network ... because every large network has P-routers;
You create BGP session templates instead of configuring two parameters of a few IBGP neighbors;
You configure MPLS in the core network instead of using BGP on all routers ... because it saves you a few BGP sessions ... and that's the way things should be done anyway;
When configuring OSPF, you define inter-AS links as passive interfaces ... not because you're running OSPF in the other AS but for security reasons :)
You compose and deploy iACLs, then CoPP policies and GTSM, all in order to remove any possibility that your XBox 360 and your MythTV DVR might gang up try and take down an eBGP session.
7. You run MPLS TE in each core between P-routers and tunnel LDP inside TE-tunnels ... because running full mesh TE between PEs doesn't scale. 8. You implement RTBH just in case 9. You set up a complex communities-driven policy, even if the goal of the lab is just to test some obscure SRC feature.
And just when you think you have got the best SP network implemented your largest customers then forces you to add a bunch of GRE and L2TPv3 tunnels to cure *their* problems.
Finally, your customer asked that they wanted your network to support X.25 switching :-)
9. You use VPLS more often then Spanning Tree.
10. You consistently bug Cisco's business units about having all devices having capable of imposing and popping tags.
8. You implement RTBH just in case
9. You set up a complex communities-driven policy, even if the goal of the lab is just to test some obscure SRC feature.
Finally, your customer asked that they wanted your network to support X.25 switching :-)