Last edit: 05-03-17 Graham Wideman

Personal

Operational Priorities

In the summer of 1996 I started to tackle SDSU's most important system, the student/course/section/scheduling/faculty system, SIMS/CDPS.  When I started work with the leads of the two SIMS/CDPS development/maintenance teams it was immediately apparent that a crisis was waiting to happen, and that this was not widely known. 

In the early '90s a plan had been established to replace the existing million-line COBOL/CICS system with one based on Oracle, to be known as SIMS-R ("relational").  The idea was that the relational environment would provide a more flexibility and capable platform for maintenance and future development.  And as a convenient side effect, this would all be done in time to avert the year-2000 problem in the existing system -- whose impact would be felt in early to mid '99.  Work on the core half of the system by the CSU central team (CASA) started in 1994, with campuses scheduled to start their efforts in mid 1995.

By mid 1996, the SIMS/CDPS analysts were very concerned because no SDSU management action had been taken to either start work on the SDSU half of the system, nor to provide the design and functional testing that CASA was relying on SDSU for...and there was no plan nor budget to start.

From that point on through 1997 I was involved in many rounds of strategizing with colleagues and inter-divisional liaisons to try to get this project attended to.   There were several significant challenges:

• Responsibility for different parts of the project were split between three different divisions, hence no clear authority for the system as a whole, and no accountability for the system as a critical campus-scope resource.
• Management in each division avoided revealing any problems with the project outside the division (and some staff were explicitly ordered not to), which made assessment and communication about the true status of the project difficult.
• There was widespread ignorance of what the year-2000 problem is, that this system had it, and thus the urgency of the Oracle-based project.
• There was no management awareness of the time and resources needed for the Oracle project they had planned to perform.

By early 97 we had secured progress on a number of critical issues:  a sizeable budget and numerous positions had been approved, the two development teams had been combined, a project space had been assigned away from the teams' normal area, and work ostensibly started in March 97.  This was already too late to provide design feedback to the central CASA, and has resulted in significant problems since. But at least our team had started.

However, there continued to be significant problems installing I.S./project/Oracle-savvy management for the project, and at this point (mid 98) the team is on their 5th management configuration.  As time progresses, there is increasing probability that the Oracle effort will have to be set aside to conduct a panic conversion on the existing million lines of COBOL/CICS code.

But of greatest importance to my task at hand, it became clear that as the SIMS-R process languished, there was diminishing possibility that any time or expertise would be available for warehouse-related activities until after the dust settled on SIMS-R, and that was likely to be around mid-2000 or later.  Needless to say, that was not encouraging. However, another group that regularly received extracts from SIMS had plans to warehouse this portion of the data, and that kept the hope alive of a nearer-term ability to bring a broad spectrum of data together into a warehouse environment.

[Back to "Strategy"]

Go to: