NOTE: I debated writing this because some might take it as criticism of Shmoocon or dismiss it as my disappointment in the failure of my plans for participation in the Shmoocon Labs this year. It is neither, it was simply a great learning experience- and like most of the best learning experiences, the lessons learned were not on the curriculum.
Shmoocon was great again this year. The Shmoobus was entertaining as always, blizzard and its aftermath on the way down, rolling LAN party and much more on the way home.
One of the things Shmoocon does every year is build their event network as a training event, Shmoocon Labs, not just to provide connectivity for the con. They swoop in, split into teams (switching, firewall, wireless, services, visualization, etc.) and start building a network on Thursday morning and by Friday afternoon they have a (mostly) functional network serving the needs of the conference and its attendees. And by Sunday evening it is all gone. The goal is to build a truly “enterprise class” network, and they pull it off every year.
The various teams handle specific segments of the network, and while everyone works together, there is a segregation of duties between teams and tasks stay in the appropriate team. Everyone on the labs crew is a volunteer, many even pay extra to participate and learn- this means expertise is based on the experience of those participating. There are a lot of very experienced network engineers in the Shmoo team, and they do a killer job…
BUT
The network is complex. It is done on purpose, as part of the training experience. But complex *anything* is problematic in many ways. Complexity brings challenges to configuration, compatibility, manageability security, and more.
AND
Task isolation between teams means that sometimes the most expedient solutions are not applied because the team with primary responsibility for the task or service is given the opportunity to work through the issue and learn.
BUT
These “problems” aren’t really problems at Shmoocon, right? They are part of the program, there are very good reasons for intentionally introducing these burdens and challenges. And besides, everything works eventually and thus the labs are a success.
Hey, wait a minute…
“That’s not a problem because we do it on purpose, and besides, it works” sounds vaguely familiar, doesn’t it? It turns out that they manage to capture more details of a true “Enterprise Network” than are in the plans.
Jack
