Yes, Im still crazy
The effort I'm going to have to put into this attempt is gi-normous. I dont know or care if it can be done, Im going to do it anyway.
The overlap of CCIE SP to R/S is quite great - I would say 60-70%. The one thing I am finding that R/S goes in so much more in depth is layer 2 technologies. SP Was all MPLS/IP with very little Layer2 of any protocol.
Working for a service provider, stuff like VTP and STP are just not used. Every link is pretty much treated as "HERE BE DRAGONS" - if a customer can screw it up - they will screw it up. Enterprise consists of more vpn and patch work, and generally broken networks. A Service Provider would be out of business if they pulled the same stuff I have seen in Enterprises.
A customer of an ISP I worked for is the perfect example. They are a very well known adult dating website. Calls kept coming in that "INTERNET SLOW. FIX. MTR SHOWS DROPS. FIX THIS RIGHT F'N NOW!!!". You can only explain so many times that a policer drop means you are sending too much data. The customer said "We do not employ network admins nor will we. It is up to you to fix it".
The customers network from what we deciphered was 100% flat, with a bunch of switches hanging off ours. It appeared that not only were there L2 loops going on, but they were doing some sort of UDP program that would blast out ignoring any sort of drops, and effectively DoSing their internet connection from policer drops on our router.
The stupid Layer2 issues that show up in enterprise networks just baffle me. STP will work if you do not mess with native vlans or allow BPDU's on access ports. I am all for keeping your network simple, but please a least get a consultant with some sort of credentials to set it up for you.
Tonight im gonna work on my LAG protocols, bpu formats and STP. This should be real fun. I would much rather be reading about much more exciting stuff, like multi-point to multi-point LSPs.