IPexpert Volume 3 Mock Lab 7 Review

This is a very challenging lab. I missed quite a few things, and there was a LOT of troubleshooting involved when things wouldn't work.

To begin with, all switches have dot1q trunk links to each other. However SW2 and SW3 are using flex-links. At first nothing seems wrong, then all of a sudden in the IGP section, SW1 becomes unbearably slow and R4 and R7 keep dropping EIGRP adjacencies. I noticed the RIP and IP INPUT processes on R1 were eating up the CPU. RIP and EIGRP packets were being looped over and over and over because STP does not run over Flex links! I shut the links down and attacked it later.

Another task asked to configure MLPPP over Frame Relay without using a multilink group. I created a multilink interface, but the answer was to use ppp multilink on a virtual-template and forget about the multilink interface.

BGP, Multicast, IP Services and IPv6 were pretty easy. I was glad because I had already spent 4+ hours getting through IGP. I did miss the HSRP task because they wanted the highest group possible. I used 255 but you were supposed to switch to version 2 and use group 4096!

Missed some delicate stuff regarding QoS. Byte counts were easy enough configure but the task said you should assume packet sizes of 100. This means you needed to adjust queue-limits also, which I did not do.

Another tricky one was a CQ to MQC conversion task. They displayed the CQ conversion as using a TCP syslog port. If you use NBAR to match syslog, it only uses UDP buy default. So you had to create a custom port-map. Tough one to see right away.

There was a login task that asked me to enable SSH for VTY lines. I forgot to create the key so it never would have worked. I should have verified this by attempting to login via SSH.

Finally, when I went back to the flex-link tasks I just used "switchport trunk allowed vlan none" to get it to work. The PG pruned even VLANs off from SW3 to SW2, then pruned the odd ones from SW2 to SW4. Then they shut the link from SW1 to SW2. Anyways, there were probably a number of ways to do it. It didn't really matter as long as you have connectivity.

Next up: Lab 8 tomorrow.