ICC Home  /  Members  /  Meetings  /  Peer Support  /  Documentation  /  Projects


ICC Notes from December 14, 2006 ITAC-NI Meeting:


done from memory--no recording--so take with a grain of salt

back to ITAC-NI notes index

    notes by Steve Lasley

    The meeting was held, as usual, on the second Thursday of the month at 10 AM in room 219 of Dauer Hall. The meeting was run by ITAC-NI chairman, John Sabin, Professor of Physics and Chemistry & Associate Dean for CLAS - Information Technology. Prior to the meeting, Chris Leopold and Dan Cromer introduced Steve Lasley to Dr. Sabin as being the third IFAS representative to the committee, replacing Joe Spooner who recently left IFAS.

    Tom Livoti, Coordinator of Management Analysis for HealthNet could not make the meeting, but was represented by Jan van der Aa. Tim Fitzpatrick (Director of CNS), Stephen Kostewicz, Shawn Lander, Dan Cromer, Erik Deumens and Chris Leopold were in attendance. Steve does not know Clint Collins, Craig Gorme or Hansford Tyler (Assistant Director, AISS at the Health Science Center) on sight yet, so they may have been there as well.

    A number of visitors were in attendance, including Marc Hoit,Russ Poole, Dan Miller, Joe Gasper, David Gagne and Dave Pokorney (briefly) among others.

    Election of Chair

    Dr. Sabin related that our committee charter calls for yearly elections for chair. After nobody offered to step into that role, John kindly agreed to continue on as chairman for another year.

    Wallplate

    Introduction

    Marc Hoit introduced the fact that the wall-plate networking services will be funded centrally. Dr. Hoit emphasized that many to most of the details are yet to work out and emphasized that this committee will be involved in approving on the implementation plan. There are two issues of a fairly non-technical nature which he asked that we carefully consider:

    • Allocation Model:
      One of the crucial matters is to decide what the basic offering will include port-wise and the charging for various services in excess of that. Obviously, unlimited ports cannot be made freely available upon request.
    • Priority of Deployment:
      The other matter relates to the rollout of this across campus. We need to consider how to handle that fairly and equitably. Naturally, each unit will have some justification why they should go first.

    Presentation of Plan

    Russ Poole began with an historical perspective. ITAC was charged by the Vice Provost back on September 21st of 2001 to:

    • Provide direction and analysis in support of a comprehensive managed approach to network infrastructure for moving voice, video, and data throughout the University Enterprise.
    • Establish Network Services to the Wall-Plate within a 3-5 year time frame

    Russ reported that, as of this November, we had implemented 10,000 ports on a voluntary basis at a cost of $5 per active port per month. Now with the Provost funding of the wall-plate we have to plan how best to implement this campus-wide. Rollout of this needs to consider our current situation and how to get where we want to go:

    • Prior to rollout of new ports, existing wall-plate projects must be completed
    • 10,000 ports currently supported
    • 1000 new in process
    • 1000 ports in queue
    • 2000 ports needing life-cycle replacement
    • Planning and staffing to implement 15,000 ports

    The processes involved in this deployment will include:

    • Inventory of current condition of the network (just finishing)
    • Documentation and labeling of physical inventory
    • Integration of inventory information into UF Facilities Planning systems
    • Physical remediation if needed
    • Network (re)-design in necessary
    • Equipment procurement as needed
    • Configuration and installation of equipment
    • Collection integration
    • Finalized Services

    Regarding the deployment rate, Russ noted that they have a track record of 10,000 ports over 2.5 years for a rate of 4000 ports per year. The challenge will be to add 15,000 ports over three years; this would mean upping the rate to 5000 ports per year. Consequently, CNS needs an addition 5-person team in addition to the two such teams already in place. The goal is to get this rollout done and then smoothly transition into support, maintenance and life-cycle replacement mode.

    This fiscal year we are looking to complete the work currently in progress as well as the queued work. At the same time we hope to finish the project planning and preparation. Next fiscal year (2007-8) we will begin deployment based on the plan which is currently being developed. The project plan will be ready for review by March of 2006 and implementation should start in July of that year.

    We still must decide on the parameters of the basic service levels. The $1.5 million/yr funding level assumes 25,000 10/1000 ports. The actual count may be higher. We need to determine:

    • What will be the basic subscription rate
      that is, the appropriate number of ports per location
    • What is the appropriate subscription rater per function?
      that is, what levels should there be for classrooms, work area, research areas, etc.
    • How do we pay for expansion of the network beyond 25,000 ports

    These considerations are still under review.

    Other Discussion of Plan

    Jan van der Aa stated that it costs HealthNet about $150/port per year. Housing reported fairly similar costs. Consequently, Jan believes the funding we have received will only cover 10,000 ports and we will have a $3 million shortfall overall. Tim Fitzpatrick noted that the costs which Jan spoke of included core network costs which are not included here; this plan is for the "edge" structure within our buildings--those switches to which the wall plates connect. There was discussion of the true costs involved and how they might be distributed variously, but the end result was that we will be short funds and must figure out ways to make that up. Of course, the concern is that we should undertake a project, especially one of this magnitude, without first getting a sustainable funding model in place.

    Chris Leopold asked if the long-term plan is to include off-campus facilities of the university. IFAS has RECs and county offices which are a concern. Dr. Hoit replied that this is certainly intended and related that he has had discussions with IFAS regarding this. Some progress is being made on hooking in RECs via the LamdaRail; county extension offices may be more problematic.

    Chris Leopold and later, Joe Gasper, asked about access to the switches by local security personnel. Dan Miller reported that they were working on a web interface which would allow the necessary access and control to security personnel.

    Joe Gasper and David Gagne asked for clarification on the WiFi coverage that would be included. Basic service includes the inside building areas. Tim Fitzpatrick noted that outdoor coverage is considered part of our core network and that no new deployment of outdoor wireless is currently being considered. In fact, maintenance of our existing outdoor wireless network is very low priority. It was mentioned that there is there is an increasing possibility that UF may be able to instituted a student technology fee; this is exactly the sort of deployment which that might be used for (as it directly supports the students).

    Erik Deumens noted that some cooperative arrangements must be made for departments that need access to switches in the support of intra-building local networking needs.

    Dan Cromer initiated a discussion on the need for QoS. Dan Miller said that providing more bandwidth is now commonly seen as the solution there. Dan mentioned that, in his experience with videoconferencing--even just here on campus, QoS may still be needed. Dan realizes that it might be an extra cost service, but was concerned that QoS be available should need dictate.

    Next Meeting

    The next regular meeting should be January 11th.


last edited 20 December 2006 by Steve Lasley