DUKE ITAC - February 10, 2000 Minutes
February 10, 2000
Attending: Ed Anapol, Pakis Bessias, John Board, Jim Coble, Nevin Fouts, Pete Goswick, Alan Halachmi, Patrick Halpin, John Harer, Alfred Trozzo, Donna Hewitt, Bob Newlin, Ken Knoerr, Roger Loyd, Melissa Mills,Caroline Nisbet, George Oberlander, Lynne O'Brien, Mike Pickett, Rafael Rodriguez, Mike Russell, Robert Wolpert, Kyle Johnson, Bob Currier
Call to Order: Meeting called to order by Mike Pickett on behalf of Robert Wolpert who is joining a little later.
Review of Minutes and Announcements:
- Concerning the agenda item dealing with mass e-mails, Mike asked for additional comments on the notes that had been sent around. He will distribute them when he has them all.
Update on Cable modem service
- the period of time to which we commit
- to what volume we commit.
Rafael Rodriquez gave an update on cable modem service offered by Time Warner. The name of the commercial product is Roadrunner. Time Warner provides this service commercially bundled with ISP service. This is in contrast with the ADSL service GTE offers through Duke which partners with Duke as ISP. Rafael is pursuing a similar arrangement with Time Warner, i.e.,implementing a Duke-specific Roadrunner service with Duke as ISP. He'd like an arrangement where Time Warner's 'Duke' package used our IP numbers, our Internet access and offered a good price. This will allow us to manage the IP #'s and handle security.
He noted that the price will depend heavily on two factors:
An advantage of Time Warner's service is that its service area includes Orange County and Wake County. Rafael is not sure about Chatham County yet. NC State and UNC have the same sorts of needs. Rafael is trying to coordinate our negotiations with theirs.
Pat Halpin asked if there is any coverage in the Marine Lab area.
Rafael wasn't sure.
George Oberlander asked about the differences between ADSL and cable modem.
Rafael answered that there are differences in the technology. ADSL is switched and cable modem is shared in a neighborhood.
Nevin Fouts, John Board and Bob Currier have both services installed in their homes.
John said that whether ADSL works to your house is black magic. It depends on the quality of the wires to the site, and differs within neighborhoods. Therefore, one can't count on it as the solution. If it works, then ADSL is a good choice. However, because of its coverage, TimeWarner's cable modem looks like it will be more scalable.
Bob Currier noted that it is true that due to the shared nature of the cable modem technology, someone in your neighborhood could take down your cable modem link.
Mike Pickett said that the cable modem technology seems to be occasionally knocked out by things like weather (maybe three times in two years, each time for almost a day). When ADSL is up, it usually stays up.
Rafael says his approach is to let the market place decide which technology to use. There are pros and cons of each. He is afraid that TimeWarner is focussed totally on the consumer end of it. He is trying to get them interested in dealing with us as a corporation.
Roger Loyd mentioned that NC Central is also a partner with us, particularly in the library world and would be a good partner to include in the Time Warner negotiations.
Rafael mentioned that a further trial may not be feasible because there are already 1,000 people in the Triangle area on the commercial Roadrunner service.
Wireless Network Access
Bob Currier Bob Currier gave an update on wireless network access. There are several products and technologies we could do as a pilot. In particular, Bob is interested in getting some things up for the students. What specific technologies should we deploy, and where should we deploy them?
Alan Halachmi noted that Duke Student Government had passed an unopposed resolution that Duke be urged to implement wireless solutions in areas like the Bryan Center Cafe and the Perkins Library 'Perk' so that students would be able to use their laptops for network access. He noted that increasingly students have laptops and would like to use them with network access when hanging out.
Melissa Mills mentioned that faculty have asked for wireless connections in classrooms.
Jim Coble said he thought the Library would be interested, though he can't speak for David Ferriero (David Ferriero subsequently confirmed that the Library was interested).
John Board has been reviewing products for a home wireless network. He found several examples of university implementations. New South Wales University checks out network cards for use in their library. The price of wireless network cards is down to $150 per card.
Nevin Fouts mentioned that roaming wireless is now available. Though it is still very low bandwidth, it is increasing. He suggested that we look not just for local wireless, but towards a solution that would allow a seamless interconnection of switches, from roaming to local as one travels. We should be ready for this when the bandwidth increases.
Mike asked for an update on standards from Bob Currier.
Bob answered that the range and penetration rate hasn't changed a lot since the Duke trial a couple of years ago. The range is five - ten miles. The access point within a building has a theoretical average of 1,500 feet, but in practice it is closer to 500 or 600 feet. The bandwidth is 1 to 11 megabytes per second per box shared. It is possible to gang boxes together to increase the bandwidth in heavy traffic areas. The biggest challenge is the roaming issue. While we can currently buy technology that supports roaming, the challenge will be to scale up to roam from point to point as the technology ramps up.
George Oberlander asked about the security implications.
Bob said listening is 128-bit DES, and is difficult to get into. Security and access are definitely issues. We will need authentication to control the situation where anyone could go to Best Buy, but a network card, come on campus and become a member of the campus network.
Pat Halpin asked if there wouldn't be applications other than wireless in the cafes. For example, Facilities staff who need to access to their information as they move from building to building and researchers in different labs.
Alan Halachmi said it was really important to provide more access to the network in places like the Bryan Center. Currently there are no clusters in the Bryan Center and only two e-mail stations which are always busy.
[Robert Wolpert arrives]
John Board asked what we can do about the authentication issue. Is it a show stopper?
Mike Russell said that Air net products are currently being used by the Med Ctr throughout the house.
Bob Currier is convening a group to review issues. Let Bob know if you're interested in joining. He'll plan to report back to ITAC in four weeks.
[Mike hands the meeting over to Robert]
Mass Electronic Communication Guidelines
Robert Wolpert introduced the agenda item concerning a policy related to use of 'bulk' e-mail distribution lists. This seems to have become a concern because the mass paper mailings used to cost money and now a mass electronic mailing is possible without charge; and because it is sometimes easy to grab a large distribution list and use it for some other, non-related purpose.
Mike Pickett said what we basically need is a group to draft language for a straw man policy for ITAC to review. George pointed out that with an e-mail distribution in the Lotus Notes server of 10,000 people, a single misuse could cost Duke $50,000 in lost wages.
Robert asked for volunteers. Caroline Nisbet, George Oberlander, Alan Halachmi and Ken Knoerr volunteered.
Pat Halpin asked for the scope.
John Board said the Steering Committee would write up the concerns in the form of a charge. Lotus Notes mailing lists and acpub mailing lists would be the focus.
Pat said the inverse of controlling was responding. Official Duke e-mail lists being sent out were labelled as spam by AOL.
Melissa said that this may be due to a misuse of the duke.edu network by others (a security issue), which caused AOL to block mail from duke.edu as spam. Charlie Register had mentioned this at the Natural Sciences Faculty IT Workshop.
Continuing Discussion of Academic IT Planning Process
- tools to yield access to materials,
- tools to create results (which includes supercomputers and related infrastructure)
- tools to disseminate results.
- executive sponsorship - is this something we need?
- sponsorship is putting money where you state the priorities are.
- endowing IT...
- the need for high quality personnel
Robert introduced John Harer and asked him to talk about IT planning for academic support.
John reported that we are at the end of Phase I of a three phase process. Phase I charged the eight schools to respond to eight questions. Also, division l faculty groups have prepared white papers. Forty-five papers were submitted in the science and engineering areas. They've found some areas where scaling resources would help many. They've been looking at the school plans. Things are in sync. Schools' plans seem to fit well together. At the Provost level, they are not prioritizing at this point. Rather, they are trying to understand and think about the implications of the plans. This part is moving well. What is not yet working well is the IT part. Now is the time to gear up here. Little has been done at the inter-school level. There are lists of problems but not a lot of solutions about how to make the investment effective. It's important to coordinate the investments and drive them to be supporting the academic mission.
There is a small IT coordinating committee - costing things out, and coordinating with Lynne's group on instruction.
John Harer said he is impressed with the list of IT concerns that Betty solicited from the ITAC group in December, 1999.
John said the biggest issue in his mind is that we've not made the case in a way such that the administration can understand what they are getting for the investment.
John noted that there is also a science and engineering working group which is different from the consideration of research computing across the range. Computing has become non- denominational (e.g., not sciences, social sciences, humanities) it is like the library now. Whose job is it to think through the fact that the economics have changed for the graduate level. Research in IT, research on IT, research using IT? Are these the same? Robert said we need to make sure that the tools are in place to be able to do our mission. John said we're doing that in the sciences, but it's not clear whether we shouldn't be doing the same thing in the Humanities. He asked if ITAC still believes in the phrase "third to none".
Nevin Fouts asked where we draw the line in the continuum of networks, security/authentication, and applications strategies.
Melissa Mills said neither a pure central nor a pure decentralized polar approach works best. A distributed approach makes appropriate use of both. Situations need to be looked at one at a time, and the model needs to be one that can change and evolve as technology evolves, and as units and their needs change.
Kyle Johnson mentioned the IT skill gap. It seems to be very far down the list of priorities, but without investing in training for how to use IT, it doesn't matter how much money one puts into IT.
George Oberlander said that research activity requires three fundamental tools:Additionally, there is an overlap among these tools: as you increase your access to materials you increase the power and scope of the tools for creating results and disseminating them. Moreover, this overlap extends to academic IT in general. Therefore, attempts to justify academic IT costs should both reflect and make use of this overlap.
Ken Knoerr mentioned the Enterprise wide systems. As more systems are rolled out,the need for end-user training becomes more pressing. He also noted the need to listen to the end-users as you develop systems.
Kyle said that as central systems are deployed, the model has been to push the support to the departments. Costs look good centrally, but costs are now at the department level.
John Harer's question is: From the academic perspective, how do we get those kind of questions answered?
Nevin said that there's not a global corporation around that wouldn't say that IT is strategically important.
As an example, John Harer mentioned the common need for materials sciences support for Physics, Chemistry, and the Medical Center. We need to find ways of generating more collaboration across the disciplines. A centralized computer lab may be the place to make that happen.