Adventures in ERP migration

August 6, 2015:  Today I had a great phone call with a former colleague in the Washington State Community College system. He joined our little band of 32 IT Directors about a year after I did, and was at a similar-sized college so naturally we hit it off.  It was great to reminisce and recall personalities, but the thing that stuck with me had to do with the Washington State ERP migration going on now, called CTCLink.  Not the technical, project management stuff — but the ways in which the migration is apparently being mismanaged and showing insensitivity to the folks who’ve run the old, outgoing system for 35 years.  In sum the new folks are unceremoniously denigrating the old folks and their life’s work, and that’s making them less willing to help in the transition.

The biggest tech problems I face are never technical, but people.

This reminded me of some things that happened at the University of Tulsa in the mid 90s when we transitioned from our Honeywell-Bull DP-6 mainframe to the new Datatel system.  For the unindoctrinated, Datatel is an Everything system for higher education — general ledger, grade reporting, financial aid, you name it.

During the process, the new dudes and dudettes were a bit disparaging of the outgoing system.  One of the SysAdmins took the last set of backups, wrapped them up tight, and took them home for storage: “In case the new system fails and we need to power up the old hardware.”  Never mind that the only other Honeywell-Bull mainframe still running in the U.S. was at Louisiana State, and they didn’t even want ours for parts…

Another highlight:  the Campus Registrar at the time purposely set up the new Chart of Accounts with 13-digit numbers.  Consultants told us: “You know, IBM with all their massive parts lists only uses 5 digits.”  We all regarded this as a not-so-subtle way to hobble the new system from the start.

The dudes and dudettes also produced T-shirts and mugs that said “Datatel is LIVE — and that’s no BULL.”

It was a very tense time.  Those of us who were not mainframe types but microcomputer and network folks were often disparaged as THEM.  There was no US — they lived in the basement of Zink Hall, and we were across campus.  They moved us all in together to try and fix the organizational culture.  Which worked only in the fact that over time ours became the dominant tribe.

One bit of capitulation that turned out to be a mistake:  Datatel was native under Unix, and other than the mainframe we were historically a DEC VAX VMS shop.  In our zeal to try and minimize the stress by offering a “familiar” platform, we chose to purchase/install the VMS version of Datatel Colleague.  It worked but not well.  In the years hence Datatel was moved to Unix where it remains to this day.

It was my first experience with a large-scale migration.  I was on the Client Services team who would sweep out and make sure all the offices were fully capable of running the new stuff — via a command line/telnet session  (Hey: it was 1994…). I learned a lot about the architecture of migrations (“cutting the hole out of the donut”) from the teams who swept in from Virginia to help us.

As an aside, our Datatel sales guy in the process was John Spear — now head of Ellucian.  A great guy to work with as I recall who did his utmost to make the migration as trouble-free as possible.  At the time, TU was one of the biggest migration clients Datatel had, and as such there was great incentive to make it work well.

Whenever I and my teams have undertaken large-scale system changes, I recount some bits of this story to remind us that aside from the cool new features, we mustn’t forget the people, with feelings and emotions that we have to migrate as well.

Leave a Reply

Your email address will not be published. Required fields are marked *