Return to
Home
Page

tired sysadmin

===================
JFI UNIX Systems Support
===================

[root@jfihome ~$] _


Welcome to JFI Computing!


We are JFI Computing, the IT support group of the James Franck Institute at the University of Chicago. We have been supporting the UNIX environment in the James Franck Institute and MRSEC going back to the 1980's, and we manage all the department and faculty-owned UNIX servers, desktops, and computational clusters. (For contact information, see our help and assistance page.)

Most of the services we provide require you to get a JFI account setup for you on our systems. This is not the same as the CNet login that the campus gives you. If you need a JFI account, please stop by our offices in Hinds Hall (HGS-033/HGS-033A), and we'll set one up for you. Accounts are only available for JFI faculty, postdocs, students, staff, or anyone affiliated in some way with at least one faculty member at the institute. Accounts are also available for visiting VIP's and scientists.


HISTORY


A long, long time ago, most of the professors in JFI/MRSEC were each running UNIX systems themselves for their own respective research groups, and giving logins to their students. (I believe this would have been in the 1980's...long before I started working here...but the stories persist.) At some point, the faculty decided that it'd be better to hire UNIX systems admins (then called "managers") to free up their own time for teaching and research, since UNIX was even more time-consuming to setup and configure in those days than it is now. (And it takes a lot of time even now!) So the first JFI UNIX managers were hired...

Other departments in the Physical Sciences Division were doing the same thing at this time. The Dean's office of the division deemed that this investment in departmental IT services was generally a good thing, and thus the Division contributed toward 50% of the salaries of each sysadmin hired by the various departments, so that the departments would each only have to pay half. An organization was formed by the newly hired IT staff called DSM ("Dedicated Systems Managers"), which allowed all of the UNIX admins in the various departments in PSD (such as JFI) to share technical advice and information. Also, a DSM listserv was created allowing us to all have access to one another's technical expertise via email whenever systems problems arose that proved to be terribly vexing. This was a resource for all the department-attached IT people in the division for many years. (The listserv was finally decommissioned...rest in peace.)

Up until recently, JFI's equipment (and even some of EFI's) was housed in our server room in RI-147 of the Research Institutes building. It resided there for over thirty years, and once included an enormous and extremely noisy mainframe, appropriately named T-Rex. As we moved away from the mainframe environment and went more and more into Sun machines, our primary server, which did almost everything on one little SPARCstation, was an UltraSPARC 1 named control.uchicago.edu. To this day, there is no shortage of people around here who remember having JFI accounts on control. T-Rex eventually went away, and as a consequence of that, some of us slowly got our hearing back. Many generations of UNIX sysadmins came and went over the years, moving on to other positions with others replacing them. There were at least eight generations of UNIX managers that I know of, and I don't even have the complete record of all of them. Most of them worked in pairs, with usually two UNIX support people working at the institute at any given time period.

In the late 2000's, the Gordon Center was built, and almost all our people moved their offices and labs over to the new building, which was connected to the RI by a 2nd story catwalk tunnel that went over 57th Street. The UNIX admins stayed, partially because there wasn't any office space left in the Gordon Center, and partially because the servers stayed too, remaining as they always had been, in RI-147. It was good for having close physical access to the machines, but our customer service suffered, because it was now our users (rather than our equipment) that we were having to "access remotely." People were less apt to informally drop by our offices with problems, or to talk about what they needed, and as a consequence, the nature of daily business became more and more server-centric, and less user-oriented.

After thirty years of continuous service, the HVAC system in the RI-147 server room was starting to become untrustworthy. Many near-disasters happened when the system unexpectedly failed, leading to system crashes, and occasionally even hardware damage because of the heat. It became imperative to find a new cooling solution, which was impossible to do in the existing room, because there was never a time when the old air conditioner could be shutdown to replace it. That would have involved shutting down every server in JFI, plus a few EFI Particle Theory Group machines and some NSIT network routers, all at the same time, for hours during the replacement. That was clearly impossible to ever arrange.

Fortunately, NSIT and the Physical Sciences Division were working on a project to consolidate all of the various departmental server rooms to save money, and a new room intended to house the equipment from all of the DSM sysadmins was being built in the sub-basement of Hinds Hall (where the "canon room" used to be!). For us in JFI though, this was more than just a financial savings -- we needed a place to move our equipment before the air conditioning systems died for good. There was no other way for our machines to survive.

Initially, the new room had air conditioning issues too, but that got worked out, and now nearly all the UNIX sysadmins in DSM have their servers relocated there. Eventually us managers were moved to Hinds Hall also.

During the machine relocation, a great deal of new rackmount server hardware was purchased, phasing out many old tired systems from RI-147 that didn't deserve a second life in the new room. Also, many of the old servers were decommissioned in physical form and replaced by "jailed" virtual servers running on the new equipment, rather than actual new machines. We are still in the process of modernizing, but a great deal of our IT infrastructure was improved during the move to HInds Hall.

We hope to continue improving our services in the future. Please let us know if there's something you think we could change that would benefit the institute.


THE SYSADMINS


Brent Busby

Brent Busby

I first discovered UNIX in 1992, when I got my account on a local "freenet," which was a public UNIX dialup system being run by Bradley University in Peoria, Illinois. These freenets were offered as a community service in many places to create a central electronic gathering place for people, and offered access to the Internet at a time when few had even heard of it outside of academic and military environments. Many other universities were offering similar freenets in other locations around the United States, Canada, and Finland, and they encouraged sharing of information between communities and academic establishments.

The one thing that struck me very quickly at that time was that the Internet was the future, and that the Internet was built on UNIX. I think I became absorbed in it far more out of personal interest than out of any intention to base a profession on it, because I didn't actually start working in the field until 1997, some five years later on.

I am also a musician. Because of this interest, I have spent a lot of time with various new wave, industrial, Berlin school, and other electronic groups of the 70's and 80's, keeping me thinking about spring reverbs, tape delays, old mixers, and all sorts of various analog noise-makers.


Tom Indelli