OARnet Update Linda Roos Gathering of State Networks

  • Slides: 26
Download presentation
OARnet Update Linda Roos Gathering of State Networks February 4, 2004

OARnet Update Linda Roos Gathering of State Networks February 4, 2004

Today’s Presentation • OARnet background • OARnet current network • Third Frontier Network

Today’s Presentation • OARnet background • OARnet current network • Third Frontier Network

OARnet Background • Founded 1987 as part of the Ohio Supercomputing Center • 90+

OARnet Background • Founded 1987 as part of the Ohio Supercomputing Center • 90+ higher ed member institutions • Board of Regents funding • OSTEER advisory council • Internet 2 gigapop

OARnet Background • Over 2 million users downstream • 1. 8 Gigabits of commodity

OARnet Background • Over 2 million users downstream • 1. 8 Gigabits of commodity internet • Ohio Valley Gigapop (Internet 2) 2. 5 Gbps • Maintain over 4000 interfaces • IP over ATM

Third Frontier Network • • Project Rationale History Backbone Acquisition and Funding Project Implementation

Third Frontier Network • • Project Rationale History Backbone Acquisition and Funding Project Implementation

Project Rationale • Started with a Bold Vision – The Ohio Plan: a $900

Project Rationale • Started with a Bold Vision – The Ohio Plan: a $900 million research and commercialization effort • Now Ohio’s $1. 6 B Third Frontier Project – A Dark fiber network was essential to the plan • Ohio has a diverse and distributed array of research universities – The state needs to use all its resources effectively • Universities need to be connected to business & industry • Goal: “researcher in Athens and colleague in Cleveland could work together as if labs next door to each others”

Rationale, continued • Network Plan was Built on a Strong History – Ohio Supercomputer

Rationale, continued • Network Plan was Built on a Strong History – Ohio Supercomputer Center and the Ohio Academic Resources Network (OARnet) were created in the mid-1980 s – School. Net created in the mid-1990 s

Rationale, continued • Why was Dark Fiber essential? – Cost and cost structure –

Rationale, continued • Why was Dark Fiber essential? – Cost and cost structure – Ability to manage the network to ensure: • Quality of Service • Smooth scalability

History • Given these circumstances, Ohio started to look at how to create a

History • Given these circumstances, Ohio started to look at how to create a dark fiber network. • In 2001, OARnet made a tentative commitment to a start-up, Mission Networks, that proposed to deploy a new fiber network to Tier 2 and 3 cities in Ohio. • In early 2002, Mission Networks was unable to complete the financing for its project and did not build the network.

History • It created the idea of a project structure where – Phase 1

History • It created the idea of a project structure where – Phase 1 would be replacing OARnet’s backbone with dark fiber – Phase 2 would connect Ohio’s public universities, private research universities, and federal labs to the network with dark fiber – Phase 3 would be connecting other universities and colleges to the network. • Recommended moving forward with an RFP for Phase 1

New Network

New Network

Backbone Acquisition • In Summer of 2002, an RFP was issued • Dark fiber

Backbone Acquisition • In Summer of 2002, an RFP was issued • Dark fiber was strongly preferred, but vendors were invited to propose alternative leased services. • Vendors who bid dark fiber were required to offer a minimum of a single pair of fiber over their network.

Backbone Acquisition, cont. • Statewide dark fiber committee decided – leased lambdas were too

Backbone Acquisition, cont. • Statewide dark fiber committee decided – leased lambdas were too expensive and not widely available. • Selected a bid from Spectrum Networks that offered a package of fiber from – American Electric Power – Williams Communications (Wiltel), and – American Fiber Systems (AFS)

Backbone Acquisition, cont. • Bid offered – Single pair of fibers – Connecting all

Backbone Acquisition, cont. • Bid offered – Single pair of fibers – Connecting all of cities where OARnet had a POP. – City rings in Cleveland Columbus • A public ring was already in place in Cincinnati – Rings offering redundant routes to major cities.

Backbone Acquisition, cont. • Fiber Purchased – More than 1600 route miles – $4.

Backbone Acquisition, cont. • Fiber Purchased – More than 1600 route miles – $4. 6 M for 20 year IRUs (Indefeasible Right of Use) – $342 K/yr for maintenance – Multiple types of fiber • SMF-28, Teralight, and Truewave Classic • Aerial and buried

Backbone Financing • Financing from Ohio State University – Long-term financing to pay for

Backbone Financing • Financing from Ohio State University – Long-term financing to pay for fiber ($7 M) • To be repaid by what OARnet had spent on its previous backbone – Short-term financing ($2 M) • To cover cost of running both networks in parallel and other start-up costs. • Financing from state capital budget – ($8. 5 M) • To pay for equipment and last-mile connections for public universities, private research universities, and federal labs (17 institutions).

Project Implementation • Guiding the project development – Multi-agency policy committee • School. Net,

Project Implementation • Guiding the project development – Multi-agency policy committee • School. Net, DAS, OBR, OSC, and OARnet – OARnet Advisory Committees • Lighting and Architecture; Price Modeling Committee; Last Mile Committee; Implementation Committee – OARnet Internal Staff Responsibilities • Project Management • Site Preparation • Network Management

Implementation, cont. • First Two Tasks: Integration and Conversion 1. Integration of purchased fibers

Implementation, cont. • First Two Tasks: Integration and Conversion 1. Integration of purchased fibers • Additional IRUs and Right of Ways are required • Fiber cannot be characterized until fibers are connected • Equipment can not be specified until fibers are characterized. 2. Preparation for Conversion by connecting new POPs to old POPs. • Reduces the time required for conversion from old network to new network • Decreases the time needed to run both networks concurrently.

Implementation, cont. • Lessons learned – With multiple types of fiber, fibers will have

Implementation, cont. • Lessons learned – With multiple types of fiber, fibers will have different core sizes and splices will create inevitable core mismatches increasing attenuation. • You want to minimize the number of these splices if you want to support high-speed bandwidth. – Integration can be as or more legally complex process than purchasing the backbone fiber. • There are many more fiber purchase, right-ofway, and maintenance agreements required.

Equipment • Architectural choices – DWDM at the transport layer • So we can

Equipment • Architectural choices – DWDM at the transport layer • So we can have multiple private networks each with its own lambda • Ability to provide Lambda for individual research project – MPLS rather than SONET • Does not use as much bandwidth for overhead • Provides equivalent fail-over times.

Equipment • Equipment Selection – Cisco selected for both Optical and Switching Gear •

Equipment • Equipment Selection – Cisco selected for both Optical and Switching Gear • Cisco 15454 DWDM • GSR 12000 Routers – Equipment bought through SBC • SBC will do fiber characterization, configure, pretest, and install equipment. • OARnet Team responsible for site preparation

Equipment • Status – Equipment staged and tested at SBC facility – Installation commencing

Equipment • Status – Equipment staged and tested at SBC facility – Installation commencing following site surveys and fiber characterization – Equipment being installed currently

Other Events • RFP has been issued for last-mile connectivity to all higher education

Other Events • RFP has been issued for last-mile connectivity to all higher education and K-12 telecom aggregation sites; vendors responded, proposals being reviewed. • Setting policies on connecting to the network – Including business use of the network.

Next Steps • Light and test backbone in April/May/June • Connect first 17 institutions

Next Steps • Light and test backbone in April/May/June • Connect first 17 institutions in June/July

Contact Information Linda Roos OARnet lroos@oar. net

Contact Information Linda Roos OARnet lroos@oar. net