History - Minutes from Washington, DC - October, 1997
From GPNWiki
Minutes of the Great Plains Network Meeting October 8, 1997
Location: Lexington Room, Hyatt Regency Hotel, Washington, D.C. Time: 8:00 p.m. - 10:00 p.m.
Attendees: Pocedes (OSU) Colley (MoRENET) Albert (OSU) Merrifield(Ark) Williams (OU) Townsend (KSU) Finkelson (UNL) Zimmerman (Ark) Mitchell (MoRENET) McCarville Summerhill (KU) Engstrom (USD) Springer (MO) Neas (NDSU) Lustgraf (ISU) Grobe (KU) Niebaum (KU)
Advisory board members: Jan Poley (ADEC) Dan van Belleghem (GU) Guy Almes (ANS and Internet2) Doug Gale (GPN Board) Tracie Monk for Peter Arzberger(SDSC)
Lee Williams reported he has talked with George Strawn concerning his participation on the GPN advisory board. Strawn feels he would have to recuse himself from decisions related to NSF grants, which could limit his effectiveness on the board. The group appears to agree that he should not serve on the board.
Niebaum presented budget information dated October 3, 1997 suggesting GPN has enough money to provide connectivity for only 1/2 of the expected time. Niebaum also proposed that all states will pay equally to cover GPN expenses.
Neas asked how NDSU Grants and Contracts group will be informed. Niebaum believes KU Grants and Contracts group will inform them. Williams suggested that KU prepare a press release to be released by all participating institutions in concert.
ISU is going to connect via MREN in Chicago. Gale pointed out that connectivity prices are not proportional to distance. Mitchell pointed out that flat fees punish those over the average costs, who tend to drop out leaving the others with higher fees and unbudgeted surprises. Summerhill argued that prices are flatter now and that the consortium will provide other services that tend to take the focus off of connectivity. Finkelson suggested that the time element involved with GPN will attenuate the danger of schools leaving the consortium.
There was some discussion of the difficulties of usage-based metrics. Neas suggested we need to report real costs rather than averaged costs on grants like those for the NFS Connections program. Gale pointed out that campus infrastructure costs reported as part of matching funds will overwhelm any other considerations.
Monk pointed out that we need to be sure we have applications in place to support future grants for continuing support.
Summerhill presented a brief history of the attack on the engineering challenge, beginning last February when all vendors wanted to provide IP bearer service. Now only IBM is willing to commit to that and it is not clear they have the understanding to actually provide services we need (c.f., IPv6).
Summerhill also proposed a network architecture developed with the GPN Engineering Committee, and would like to produce a Request for Proposal (RFP) within a week. Proceeding with an RFP would require direction from the participants.
Monk pointed out that caching EROS data in KC could protect the DS3 linking KC and EROS, and that we should make any caching effort part of the NLANR cache hierarchy. Almes suggests we look at funding help from EROS if we end up providing access to EROS data from the rest of the world.
Niebaum asked how switches in KC and EROS will be funded. Summerhill replied that the group will have to come up with funds for them. Cisco will provide grants for Internet 2 schools, but has no current plans to provide funds to GigaPOPs. Circuits overwhelm equipment costs. Almes suggested considering alternative media such as frame relay.
Summerhill reported that all vendors inquire "Who are you?" and "Who will be paying the bill?" and suggested the creation of an organization to transact the business of building the network. Finkelson pointed out that we can't move on vBNS connectivity within the context of the EPSCoR grant. Summerhill added that vendors also ask "What are your goals?" Neas pointed out that vBNS was a stated goal of the EPSCoR grant.
Niebaum wanted to buy an IP bearer service from the beginning for a variety of reasons, but now believes we need to construct our own network and will require an organization to do that.
Presumably we would have an organization to which we would pay dues....and those dues would be around 30K$ per month plus 10K$ per month if we share commodity Internet connectivity (Engstrom). I1 connectivity is particularly attractive in light of the recent announcement by Global Internet that they intend to sell their network business (which currently provides I1 connectivity to 5 of the GPN state networks).
Finkelson suggests that any organization would also need staff, even if some costs were reduced by member institution largesse. Poley has experience with non-profits and warned against underestimating management costs and effort. She recommended putting money values on any donated efforts.
Summerhill would like to make a statement of intent to form an organization part of the RFP. Albert directed us to "Make a business plan. We all have good intentions here, and the road to ..."
Poley broached the idea of using ADEC as an umbrella organization. ADEC includes 50 land grant institutions; all GPN members are also members of ADEC. Albert reported that MIDnet is still alive and is a 501C3.
Mitchell asked "Can we pass the hat to build a business plan." Albert agreed to "advise" Finkelson and Summerhill in that endeavor.
Poley suggested we use the MIDnet shell for purposes of fulfilling the EPSCoR project (GPN), while putting together an organization. The Center for Research, Inc. (CRINC) at KU is 501C3 and would probably provide an adequate platform upon which to put together the GPN.
Minutes by Michael Grobe (KU)
Caveat: These minutes have not yet been approved by the assembled group.
Attendees of the October 8, 1997 GPN meeting in Washington, D.C.:
Minutes for the October 8, 1997 meeting were distributed a few days ago. One attendee was inadvertantly omitted from the list of attendees included in the Minutes. As a result, I will be issuing an updated version of the Minutes and would like to rectify any other mistakes or omissions at the same time.
If you have noticed other errors please respond to this list or to me at grobe@ukans.edu.
Thanks,
michael grobe grobe@ukans.edu academic computing services the university of kansas
Hello:
Here is an updated version of Minutes of the GPN meeting held on October 8, 1997 in Washington, D.C.
This update adds one attendee's name, omitted inadvertantly from the previous version. No other suggestions were received in response to an e-mail request for other corrections.
- Michael Grobe
Minutes of the Great Plains Network Meeting October 8, 1997
Location: Lexington Room, Hyatt Regency Hotel, Washington, D.C. Time: 8:00 p.m. - 10:00 p.m.
Attendees: Pocedes (OSU) Colley (MoRENET) Albert (OSU) Merrifield(Ark) Williams (OU) Townsend (KSU) Finkelson (UNL) Zimmerman (Ark) Mitchell (MoRENET) McCarville Summerhill (KU) Engstrom (USD) Springer (MO) Neas (NDSU) Lustgraf (ISU) Grobe (KU) Niebaum (KU) Weir (UNEB)
Advisory board members: Jan Poley (ADEC) Dan van Belleghem (GU) Guy Almes (ANS and Internet2) Doug Gale (GPN Board) Tracie Monk for Peter Arzberger(SDSC)
Lee Williams reported he has talked with George Strawn concerning his participation on the GPN advisory board. Strawn feels he would have to recuse himself from decisions related to NSF grants, which could limit his effectiveness on the board. The group appears to agree that he should not serve on the board.
Niebaum presented budget information dated October 3, 1997 suggesting GPN has enough money to provide connectivity for only 1/2 of the expected time. Niebaum also proposed that all states will pay equally to cover GPN expenses.
Neas asked how NDSU Grants and Contracts group will be informed. Niebaum believes KU Grants and Contracts group will inform them. Williams suggested that KU prepare a press release to be released by all participating institutions in concert.
ISU is going to connect via MREN in Chicago. Gale pointed out that connectivity prices are not proportional to distance. Mitchell pointed out that flat fees punish those over the average costs, who tend to drop out leaving the others with higher fees and unbudgeted surprises. Summerhill argued that prices are flatter now and that the consortium will provide other services that tend to take the focus off of connectivity. Finkelson suggested that the time element involved with GPN will attenuate the danger of schools leaving the consortium.
There was some discussion of the difficulties of usage-based metrics. Neas suggested we need to report real costs rather than averaged costs on grants like those for the NFS Connections program. Gale pointed out that campus infrastructure costs reported as part of matching funds will overwhelm any other considerations.
Monk pointed out that we need to be sure we have applications in place to support future grants for continuing support.
Summerhill presented a brief history of the attack on the engineering challenge, beginning last February when all vendors wanted to provide IP bearer service. Now only IBM is willing to commit to that and it is not clear they have the understanding to actually provide services we need (c.f., IPv6).
Summerhill also proposed a network architecture developed with the GPN Engineering Committee, and would like to produce a Request for Proposal (RFP) within a week. Proceeding with an RFP would require direction from the participants.
Monk pointed out that caching EROS data in KC could protect the DS3 linking KC and EROS, and that we should make any caching effort part of the NLANR cache hierarchy. Almes suggests we look at funding help from EROS if we end up providing access to EROS data from the rest of the world.
Niebaum asked how switches in KC and EROS will be funded. Summerhill replied that the group will have to come up with funds for them. Cisco will provide grants for Internet 2 schools, but has no current plans to provide funds to GigaPOPs. Circuits overwhelm equipment costs. Almes suggested considering alternative media such as frame relay.
Summerhill reported that all vendors inquire "Who are you?" and "Who will be paying the bill?" and suggested the creation of an organization to transact the business of building the network. Finkelson pointed out that we can't move on vBNS connectivity within the context of the EPSCoR grant. Summerhill added that vendors also ask "What are your goals?" Neas pointed out that vBNS was a stated goal of the EPSCoR grant.
Niebaum wanted to buy an IP bearer service from the beginning for a variety of reasons, but now believes we need to construct our own network and will require an organization to do that.
Presumably we would have an organization to which we would pay dues....and those dues would be around 30K$ per month plus 10K$ per month if we share commodity Internet connectivity (Engstrom). I1 connectivity is particularly attractive in light of the recent announcement by Global Internet that they intend to sell their network business (which currently provides I1 connectivity to 5 of the GPN state networks).
Finkelson suggests that any organization would also need staff, even if some costs were reduced by member institution largesse. Poley has experience with non-profits and warned against underestimating management costs and effort. She recommended putting money values on any donated efforts.
Summerhill would like to make a statement of intent to form an organization part of the RFP. Albert directed us to "Make a business plan. We all have good intentions here, and the road to ..."
Poley broached the idea of using ADEC as an umbrella organization. ADEC includes 50 land grant institutions; all GPN members are also members of ADEC. Albert reported that MIDnet is still alive and is a 501C3.
Mitchell asked "Can we pass the hat to build a business plan." Albert agreed to "advise" Finkelson and Summerhill in that endeavor.
Poley suggested we use the MIDnet shell for purposes of fulfilling the EPSCoR project (GPN), while putting together an organization. The Center for Research, Inc. (CRINC) at KU is 501C3 and would probably provide an adequate platform upon which to put together the GPN.
Minutes by Michael Grobe (KU)
Caveat: These minutes have not yet been approved by the assembled group, although NO suggestions for change were received in response to an e-mail query.
