I've been a sysadmin of one sort or another for quite a while now, and have been doing it full-time since 1996. I've worked in various places, including:

• The Department of Physics and Astronomy at the University of Rochester, where I also graduated with a history degree.
• The San Diego Supercomputer Center in both their IT Systems and High Performance Computing groups.
• The Cooperative Association for Internet Data Analysis (CAIDA), a research group that does a lot of work measuring networks (among lots of other things).
• The Centro Svizzero di Calcolo Scientifico / Swiss National Supercomputing Centre in Ticino, Switzerland working on the LHC LCG and administering other HPC systems.
• The Administrative Computing and Telecommunications department at the University of California, San Diego (UCSD), where I'm doing mostly email, active directory, and virtualization.

Many years ago at SDSC I did some early work in low-cost commodity storage of the sort that has now become ubiquitous. It became the subject of discussion recently, and I managed to dig out the web page about our work in building a 1TB file server for less than $5000.

This work resulted in my only formal publication: Data Grids, Collections and Grid Bricks, Arcot Rajasekar, Michael Wan, Reagan Moore, George Kremenek, and Tom Guptill, 20th IEEE/ 11th NASA Goddard Conference on Mass Storage Systems and Technologies (MSST2003) San Diego, California, April 7-10, 2003.

More recently, I did a bit of quick math on this google spreadsheet showing that you can now do a 100tb file server for the same amount of money.