C2IG Meeting Minutes October 1, 1998

Present: Larry Murdock, Kevin Schmidt, Stephen Pope, Rich Prohaska, Jason Simpson, and Elise Meyer

RFP Evaluation Criteria Point Assignments

Most of the discussion revolved around how the points were calculated and combined, rather than on the actual point values. The point was raised that, since the evaluators' scores were being averaged rather than normalized, an evaluator scoring vendors higher would have a bigger impact than the evaluator scoring vendors lower. We discussed the fact that the evaluators were given the final scores and if they felt there was a problem with how the scores were combined they could discuss how to rectify the problem. Those present decided to use the same evaluation criteria weighting as the first RFP. I said that I would poll the non-present members to see if they had any disagreement (which I've done via phone calls and a separate message sent out yesterday).

Fiber Status

1. Qwest's connection to campus

Qwest has successfully spliced onto campus fiber near where Los Carneros crosses the railroad tracks. Some preliminary tests were done, and we were waiting on an attenuation cable to connect the patch panel in Public Safety to the Cisco Lightstream 1010. (Note: From Rich's mail message 10/2/98, there is now additional equipment needed at both the Lightstream and at Quest's end which will delay the final connection for another couple of weeks.)

2. On campus fiber projects

Rich said that David Chapman would prefer to schedule the Ring 2 Split and Ring 4 Completion projects to maximize contractor labor. This means that they would both finish up around the same time i.e., end of November or early December, instead of one fin ishing before the other. Elise noted that that sounded okay given the current schedule, but if we only received a few RFP responses, then we might be done sooner. Digging for these projects is supposed to commence Monday, October 5th.

CENIC TAC Update

1. UCNet and CalREN-2 merger

UCOP had planned to move UCNet over to CalREN-2 as soon as it was stable. There had been talk that this could occur as soon as January 1, 1999. At the last CENIC-TAC meeting it was announced that UCNet was going to instead replace the current SMDS lines with T3 lines which would be cost effective, and the date of merging UCNet and CalREN-2 was put off to some unknown date in the future.

2. CalREN-2 is a research network

The CENIC TAC also discussed that although some campuses were putting production traffic onto CalREN-2, it was still really just a research network. They were also concerned with the capabilities of the Cisco GSR's and they were wondering if they should implement an OC-3 "undernet" for more reliable transit. There was discussion that this didn't really make sense to us.

3. Campuses using a similar design

Stanford and maybe two other campuses were using a design similar to ours. Other campuses were dividing the bandwidth up at the Cisco Lightstream.

GSR Status

Here is the current schedule for Cisco having the features that we need in the GSR:

1. Access Lists (ACLs) - November
2. Gigabit Ethernet - December

Implementation To Do List

1. Infrastructure tasks
1. Qwest to Public Safety Lightstream 1010
Status:
1. Physical: Fiber is spliced, but additional equipment is needed at each end.
2. Logical: Once physical is completed, it can begin.
2. PS Lightstream 1010 to PS GSR

Status: Rich is staying on top of the GSR's progress with Gigabit Ethernet.

3. PS GSR connection over singlemode fiber to North Hall BTP
1. SMF fiber has been terminated for this connection. - DONE
2. Jumpering to make connection.
4. NH equipment site identification & preparation:
1. Do we need to authorize any funds to get estimates for this work?
2. Identify location.
3. Get permission to use location.
4. Identify existing rack space or purchase new rack
5. Decide UPS minimum runtime requirements.
6. Identify existing uninterruptible power source or purchase UPS.
7. The NH machine room has adequate security and is accessible to NOC staff.
8. Is any additional cooling needed?
5. Fiber required to run from NH BTP to NH equipment location (hand-off panel).
1. Do we need to authorize any funds to get estimates for this work?
2. Determine what combination of SMF and MMF is adequate for backbone and research group connections.
   Need at least:
   1. SMF connection to PS
   2. SMF connection to Eng I
   3. MMF - 4 research group connections
   Possibly need:
   1. SMF to Gigabit Research Groups, only possible HEP and Arts.
6. NH equipment - per RFP SX-3161
7. NH BTP single mode connection to Eng I BTP in room 1175
1. Do we need to authorize any funds to get estimates for this work?
2. What tasks need to be identified that are not included in the Ring 2 split project?
3. SMF fiber jumpering in Broida.
8. Eng I equipment site identification & preparation:
1. Do we need to authorize any funds to get estimates for this work?
2. Identify location. - 3152C
3. Get permission to use location.
4. Identify existing rack space or purchase new rack
5. Decide UPS minimum runtime requirements.
6. Identify existing uninterruptible power source or purchase UPS.
7. The Eng I machine room has adequate security, arrangements need to be made so that it is accessible to all NOC staff.
8. Is any additional cooling needed?
9. Fiber required to run from Eng I BTP to Eng I equipment location (hand-off panel).
1. Do we need to authorize any funds to get estimates for this work?
2. Determine what combination of SMF and MMF is adequate for backbone and research group connections.
   Need at least:
   1. SMF connection to NH
   2. MMF - 1 research group connections (assume 3 Eng I research groups already have a fiber presence in 3152C)
10. Eng I equipment - per RFP SX-3161
11. Research Group Connections to switches:
1. Do we need to authorize any funds to get estimates for this work?
2. Do we need to fill out fiber request forms to allocate the campus fiber between the research groups and NH or Eng1?
3. Fiber jumpering and testing from the research group BTPs to the NH or E1 BTP and then on to the equipment location.
4. Determine distance and db loss from research group to equipment location in order to know possibility of Gigabit Ethernet. Gigabit Ethernet constraints:
   1. Cost of backbone switch i/f
   2. Distance Limitations: MMF-SX 220-275 meters, MMF-LX 550 meters
5. Identify BTP closest to research group router or switch location.
6. Identify any building wiring that needs to be done to get from router/switch to BTP.
7. NH Research Group Connections:
   1. HEP Broida to NH via Ring 2 split (1 building away)
   2. Arts to NH via new Ring 4
   3. Alexandria: Library to NH via Ring X Alexandria: Girvetz to Library via existing FOIRL attachment
   4. ICESS: Ellison to NH via Ring 2 (2 buildings away)
8. Eng I Research Group Connections:
   1. Alexandria: Eng I
   2. ITP: Kohn to Eng1 via ring 2 (1 building away)
   3. SuperWeb: Eng I
   4. Multimedia: Eng I
9. Public Safety Research Group Connections: (if required)
   1. Is there adequate terminated MMF between NH & PS for this?
   2. Music to North Hall via Ring 2 (2 buildings away) to Public Safety

2. Other issues that must be addressed before our recommendation can be written:
1. Network Budget Issues
1. Management software costs - from RFP responses
2. Maintenance Costs - from RFP responses
3. Operating Budget
4. Staffing Costs
5. Installation Costs
2. Who will operate the CalREN-2 Backbone?
1. Installation
2. Operation
3. Maintenance
3. Must determine operating expectations, e.g., 7 by 24 operation implies more staff support
4. Policies
1. Usage and/or bandwidth
2. VLAN implementation and management
3. Class of Service
4. Identify any requirements for research group connection, e.g., if you want to connect via an 802.1Q trunk port, then your equipment must support it.
5. Spontaneous VLAN membership
6. New connections
7. Identify issues that we can anticipate, but might be better handled by the BEG.
8. Research and/or Production network

Disposition of Lists

It was decided that, once the group had a chance to review the to-do lists, the fiber infrastructure list would be forwarded to Communication Services for their comment, and the first thing to be worked on would be to finalize the backbone switch equipment locations. Paul Valenzuela had been identified as the best person to pursue the North Hall location, and Kevin Schmidt was going to pursue the Engineering I location.

Back to Main CalREN-2 Page

EMM