February 21, 2008

To: UC Davis community

Re: Email Review Committee Report and Preliminary IET Management Response

This past fall, after a major disruption of the new campus email service, I asked the dean of Engineering, Enrique Lavernia, if he would lead a formal review of the disruption, the technical planning for the service, and the communications effort during and after the incident. I am delighted at the outstanding work that Dean Lavernia's Email Review Committee performed on behalf of the campus and I want to confirm my full endorsement of its recommendations. I also want to thank the committee members (see the charge letter) for taking time from their many priorities to perform this service and, as well, express my appreciation to IET staff for their responsiveness and contributions to the committee's review.

On behalf of IET and the Email Review Committee, I am pleased to share two key documents: the Email Review Committee's Report and the accompanying preliminary management response from IET. The email review report is extraordinarily complete and thoughtful. Reflecting the careful work of a small group of faculty experts and senior campus technical leads, it presents findings in three key areas - Planning and Execution, Technical Considerations, and Communications - and makes specific recommendations in each section. The preliminary management response from IET documents work already underway or planned for the immediate future, with specific deadlines, to address the report's recommendations.

The report is the outcome of a valuable experience and process for our campus. It represents both a critical guidepost for improving IET services and a "best practice" for transparent reviews of complex issues that may be of use to other groups on campus. While many may not be interested in the specific findings, all might appreciate the importance of the process.

To fully appreciate the outcome of this process, a bit of background might be helpful. As reflected quite accurately in the report itself, when the review began the cause of the week-long, early September campus email failure was unknown, and Data Center staff were keeping the system running with admittedly interim, stop-gap measures. The review was designed in part to closely examine aspects of planning, technical design and implementation in hopes of finding a weak link that could have resulted in the system's failure and continuing performance problems. Before the review was completed, we learned that the weak link was actually a "bug" in the operating system software provided by Sun Microsystems, a diagnosis that for some weeks eluded both the international community of developers of the software (non-commercial software called Cyrus) and the highest levels of Sun technical support. Some might underestimate the value of those parts of the review that we now know had no relationship to the failure. In reality, however, the recommendations will help structure a series of steps that will strengthen our approach to selecting, implementing and maintaining enterprise-level IT applications far into the future.

The value of the dean-led review extends beyond the specific email failure in at least two important ways. First, it suggests that serious effort will be needed from vendors such as Sun (and from their customers) to ensure responsive and timely support for systems - specifically related to the increasingly common non-commercial, community-developed software, often termed community-source. And second, the review underscores the significant benefits that can be realized from well-led, collaborative efforts between administrative units and faculty subject matter experts. While the attached review clearly documents these benefits in a reactive, "post-event" setting, there is every indication that corresponding, and perhaps even greater, benefits can be realized by integrating many of these steps early on in the planning and implementation phases.

I and the IET community look forward to working closely with our campus colleagues as we implement the action steps outlined as a result of this review.

Sincerely,

Peter M. Siegel
Vice Provost for Information and Educational Technology
and Chief Information Officer