Bridging the gap between CCIE RS and SP

Just a short post regarding the above, should have posted this long ago.

I often get the question or see someone asking this on the forums. “How much switching is required in the SP lab?” or “Can I prepare using only dynamips?”. The short answer is you can rely 100% on dynamips for SP preparation. If you are not comfortable with just using dynamips, you can use the breakout switch method. Although this is overkill for SP it is a nice setup for RS. I ditched the real switches in favor for the dynamips switch modules due to the noise factor.

Another thing to note is that L2 switching is pretty much completely preconfigured. See page 11 of the attached Preparing-SP-lab-part1 22Dec08 d-3927.pdf on the cisco learning network page https://learningnetwork.cisco.com/docs/DOC-3927 to get further details.

I started full scale labs with the IPX vol 3 labs 1 – 4. I am roughly following Ed and Zay’s recommendation, which is to do IPX vol 3 labs before INE vol 2 labs and then lastly do IPX vol 2. This is because the IPX vol3 is actually just IPX vol1 labs 26 – 30 branded as IPX vol3. Therefore they fit like a glove after completing IPX vol 1:

IPX v3 Lab 1: If you are preparing for the new version 4 RS lab, this lab will be perfect for the new RS blueprint. This is a real easy lab, so get it done early in preparation. For the SP labs it is a good warm up.

IPX v3 Lab 2 – 4: Labs get progressively harder. While lab 1 took 4 hours to complete, lab 4 took 10+ to thoroughly work through.

IPX v3 Lab 5: I skipped this lab according to the recommendation. I will do this lab somewhere at the end. Therefore I can not yet comment on this lab.

Overall labs 1 – 4 are good intermediate labs. I definitely agree with Zay and Ed that these are the labs to start with when moving over to full scale labs. Most of the sections could be done on dynamips, except for things like turning off DTP and switchport security. I have not felt a real need for real switches during these labs. Lab 4 has a qinq scenario linked to a l2tpv3 tunnel that I could not quite do but worked around the qinq limitation by configuring a trunk and l2tp tunnels for each vlan.

I hope to do all the labs at least twice before September, but on the second run I will probably skip lab 1. Not much of a review, hope to add some meaningful insights in round 2

I managed to complete volume 1 of the Ipexpert SP series. The workbook consists of 25 sections of which I were able to do 24 on dynamips. The other section I could not do due to my dynamips labs still running the 16 port module at that point. It is a minor as it was basically some vlans, spanning tree and port channels.

My first impression was that the book will be full of mistakes and generally rough around the edges. After the first few labs it was not all that bad. To be honest the labs are quite good. What I like about these Volume 1 scenarios is that the ip addressing is the same throughout the workbook. This make it easier to get to the meat of the section without wasting too much time learning the topology. The proctor guide is really good as well. Some more verification and explanation wont hurt.

On the negative side the dynamips files supplied by IPexpert dont work for this workbook. A lot of time was spent creating a working dynagen .net file. There are also no initial configs for the dynamips topology meaning before every lab you have to create initial configs. This chews away at your quantity:quality ratio. I will explain: This workbook took 50hours to complete. Had the initial configs been available it would have taken 40hours.

To summarize: after the first two or three labs I had very low expectations, but later on the quality really improved and I plan to do this workbook again before starting full labs.

Oh my word this application is so frustrating. A week into prep and I am still stuffing around with the emulator, begging it to do what I want it to do. The thought of actually giving up on SP due to the issues have crossed my mind. The gns3 files provided by Ipexpert is just not on par and riddled with mistakes. If you are planning on using dynamips for Ipexpert studies, keep an eye on the blog as I will post my dynagen files when im confident everything works.  The more sections I complete the more mistakes are ironed out of the dynagen files, therefore soon I will have a working setup. Still awaiting delivery of more 3550’s switches. I want to try the alternative method of bridging out dynamips to real switches. I will attempt to trunk between dynamips and 2950’s and then patch the access ports from the 2950’s into the 3550’s. The multiple usb ethernet adapters have been a bit of a mess, so lets see where this takes me. Worse case will be q-in-q tunneling will not work, but Im not really concerned about loosing this functionality. At the moment Im more concerned with getting a functional and stable dynamips setup.

Update: A year later… well I got used the beast and we became friends. I rebuilt the configs and posted them on http://21500.net/?p=958 about a year to late.