IPv6 EIGRP NTP

Well, the configuration did not go as smooth as I expected , encountered several issues , eg like some interfaces shutting down. The problem I have to research on is configuring NTP for IPv6  seems to be a little tricky requires ntp version. ????

IPv6 Bandwidth effects (EIGRP)

I have noted that different versions or setings of routers have a certain bandwidth level cut off where when reached routes will be changed.That's routes will be selected or changed when they drop below these levels. So far I have noted 3 different cutoff levels 64kbps available bandwidth of 37kbps ; 1180kbp available bandwidth of 885kbps and recently 125ks available bandwidth of 93kbps. This raises a question , is it because of the version of IOS or is it anything to do with router settings? This may need further investigations.

Running IPv6 networks

Today I will be running tests on test scenarios -IPv6 network. I will test effects of load and bandwidth on convergence speed.

Does Md5 affect convergence?

The ping tests were done on the same topology used on (14/07/12) my last post.

Md5 Authentication on Routers in same EIGRP AS seems not to change the speed  of convergence. As revealed by the ping results above and wireshark results. There’s need to check whether it does on the other routing protocols. I think it does on RIP. 

Can you guys run the tests on the topology attached previously.

Wireshark was enabled to capture results on R2 interface S1/1 and R4 int S1/0 .

200 Pings were being sent from R5 to Lo1 address 1.1.1.1 of R1

Interface S1/0 of R2 would be shut during the pings.

This should be repeated 10 times and all results captured.

I will wait to hear your findings as well.

Thanks

This is the topology which I have used for the following tests.

I have found out that the convergence time of the routers in the above  topology can be calculated from the number of pings lost multiplied by 2 which is the specified  timeout. In the first output we lost 8pings x 2 =16seconds. I verified this also from the wireshark output which gave me  15.8840seconds. It took about 16seconds after shutting down R2’s serial interface S1/2 for the routers to notice that and begin to use an alternate route to R1’s loopback network 1.1.1.1 via R3. (R5àR4àR3àR2àR1)

Bandwidth

I have noted that bandwidth would not affect path selection until the available bandwidth on the preferred route falls below 885kps. If it does not fall below that the preferred route would remain in use regardless of the huge differences say 7500kbps vs 885kbps.

From the above outputs does not show any significant differences on the speed of convergence at the moment. It needs to be tested with traffic flowing in the network to figure out if the results will be different.

Sorry I have had difficulties in posting all my test result I will put them on google docs.

Progress

Well, I have been doing tests on the first/initial topology for the last two days. I will be posting the results, am currently analyzing the results so that I can put the results together with brief observations. I have been trying to rectify our last shortfall on our last capture of results- I have greatly improved that. I will also run the other routing protocols to capture the results for analysis. Another disappointing issue is access to the lab, losing a lot of time waiting for security to open up lab- cards no working. I have experienced that today and Anthony was letting me in during the day for last two days (mon and Tues)

Yesterday work progress

Hi Egious we finally fixed and configured topology 3 right after you left but we didn't have enough time to do more testing as they lock down the whole 3rd floor last night.

Tevita