CLAC Conference 2006 Hamilton College June 15, 2006 - PowerPoint PPT Presentation

About This Presentation
Title:

CLAC Conference 2006 Hamilton College June 15, 2006

Description:

Copper in buildings varied between Cat 3, Cat 5 and Cat 6E (by building) ... Commercial VoIP (e.g. Vonage or Skype) Voice over Wireless LAN (VoWiFi) Design Goals ... – PowerPoint PPT presentation

Number of Views:56
Avg rating:3.0/5.0
Slides: 32
Provided by: markj7
Learn more at: https://www.hamilton.edu
Category:

less

Transcript and Presenter's Notes

Title: CLAC Conference 2006 Hamilton College June 15, 2006


1
CLAC Conference 2006Hamilton CollegeJune 15,
2006
A Change of Voice Considerations in Implementing
IP Telephony at Swarthmore
Presented by Mark Dumic Associate Director
Networking Telecommunications Swarthmore College
2
Swarthmore College
  • Suburban Philadelphia
  • Highly selective liberal arts college
  • 1,450 FT undergraduate students
  • 600 Faculty-Staff
  • Students from 50 states and 46 countries

3
About the Project
  • End of Life of Intecom E
  • Infrastructure issues
  • Pure fiber distribution between buildings no
    monolithic copper plant
  • Copper in buildings varied between Cat 3, Cat 5
    and Cat 6E (by building)
  • High awareness of lack of disaster recovery with
    current system

4
Previous Student Telephone Environment
  • Digital phone per student, separate DID and VM
    accounts
  • Eliminated LD after Spring 2004
  • No student technology fee or monthly phone fee
  • College philosophy toward students
  • 90 cell phone penetration
  • Declining use of VM by students

5
College Decision Process
  • PBX Committee Members
  • Director of Information Technology Services
  • Assoc Dir Networking, Systems Telecom
  • VP Finance
  • Controller
  • Dean of Students
  • Director of Housing
  • Provost
  • Facilities Technology Coordinator

6
College Decision Process (cont.)
  • Student venues and what we heard
  • Dont want to give out cell phone numbers
  • 10-15 of students dont want or cant afford
    cell phones
  • International students
  • Cell phone reception problems
  • DONT TAKE OUR WIRED PHONE AWAY
  • No clear consensus about student services or
    technology throughout process
  • Cost a concern (for College and students)

7
Get Help!
  • Retained Compass Consulting International, Inc.
    (now Acentech, Inc.)
  • Assist with defining investigating options for
    students
  • Evaluate issues associated with technology
    options (IP telephony)
  • RFP and Implementation

8
Evaluation of Student Technology Options
  • Digital phones in rooms
  • Analog phones in rooms
  • IP PBX hard or soft IP phones in rooms
  • Outsourced PBX for residence halls
  • Traditional Centrex
  • IP Centrex
  • College-issued cellular phones
  • Student-selected cellular phones
  • Commercial VoIP (e.g. Vonage or Skype)
  • Voice over Wireless LAN (VoWiFi)

9
Design Goals
  • PBX Technology IP, TDM, hybrid
  • Analog, digital, IP, softphones...
  • No killer app
  • Decided to not specify PBX technology in RPF -
    possible cost savings in TDM or hybrid?

10
Design Goals (cont.)
  • Preserve future flexibility
  • Choice of options for maintenance, parts,
    augmented staffing
  • Financially stable company with strong US market
    presence
  • More control over upgrades
  • Minimize vendor lock-in
  • Maximize useful life
  • Adapt to future applications standards (SIP)

11
Design Goals (cont.)
  • Maintenance Considerations
  • Lots of demands on small staff - continue
    outsourced support
  • Telco Agency - no finger-pointing
  • Swarthmore is high touch
  • High vendor contact communication
  • Like close working partnerships
  • Big fish in a small pond
  • Plan tailored to take advantage of what we do
    best
  • Flexibility to have vendor do what we cant do
    well
  • These considerations favored local support rather
    than National support model

12
Design Goals (cont.)
  • Separate Voice and Data Networks
  • (Initial Scenario)
  • Wanted telecom vendor to have complete
    responsibility beyond MACs
  • Small internal staff needed for other areas
  • Staff had little VoIP expertise
  • But envisioned combining networks later
  • Recently purchased switches not powered
  • (not much incremental cost to keep separate)

13
RFP Process
  • Design
  • Didnt dictate IP or hybrid
  • Did decide on analog student lines
  • Cisco network for future integration
  • Required 3 student scenarios
  • Explored boosting of cell phone coverage
  • Pre-screen
  • RFP
  • Bid process

14
RFP Process Responses
  • Responses
  • Evaluation results
  • Proposal review
  • Financial analysis
  • Finalist meetings

15
RFP Process Selection
  • Selected Cisco Call Manager from TransNet
    Corporation in Somerville, NJ
  • Decision based on the capabilities of and
    chemistry with the reseller
  • Comfort level with converged network
  • 24x7 monitoring problem resolution for data
    network for cost of just PBX maintenance
  • Enhanced reliability technical depth for
    network
  • Significant Higher Education experience
  • Ability willingness to dovetail support with
    current staff, telco management

16
RFP Process Selection (cont.)
  • Major Benefits of Cisco, too
  • Long term presence in US telecom market
  • Familiar technology to Swarthmore
  • Excellent competitive market for maintenance,
    components, staffing, 3rd party apps
  • Good disaster recovery architecture
  • The most future-proof system proposed
  • Best E-911 and malicious call trace capabilities

17
New System Details
  • Decided on a converged network
  • Totally redundant 6513 cores
  • 3750s in-line power in buildings with links to
    each core
  • Redundant Call Managers Unity (VM) servers
  • 1,100 IP phones (faculty/staff)
  • 300 analog phones (public, FAX, etc.)
  • 1,300 analog lines (one per student)

18
Cisco 7941g Phone
19
Student Reaction
  • I think this is the best of possible outcomes.
    Im glad that this system was within our budget
    and I think the administration has dealt with
    this in a responsible manner.
  • Student Council Co-President
  • I think this is basically a pretty sweet deal
  • Associate Dean of Student Life

20
Blended Support Model
  • Initial Assumption was separate vertical networks
    for voice data
  • Enhanced support at minimal cost by utilizing
    horizontal support split
  • Monitoring of entire network
  • Quick access to large staff of network engineers
    for problem resolution
  • College boots on the ground for closet issues
  • Self-spare 3750s VG224 saved

21
Flexible Service Agreement (FSA)
  • New territory
  • Needed outside expertise/support
  • Maximize in-house capabilities
  • Didnt want to pay money for unused hours
  • Took time effort to craft the agreement

22
FSA Structure
  • Monitoring
  • annual flat fee
  • 24x7
  • Remote VPN and phone support
  • pool of hours
  • expire annually
  • On-site
  • pre-paid account
  • different rates for different times/service
    levels/expertise
  • roll-over to subsequent years

23
Convergence Saving at a Small Institution
  • Compelling Cost Reductions led to successive
    network integration
  • completely separate voice network
  • combine core and switch stacks but keep separate
    voice data jacks
  • Also use single jack for phone computer (reduce
    switches required)

24
Convergence Cost Reductions
25
Significant Findings (cont.)
  • Network architecture varies
  • Routed segments or flat L2
  • Redundancy core, servers, links
  • Firewalls
  • Vendor support for Cat 3 varies
  • Not many support Cat 3 (even just for phone)
  • If testing or certification is required, when is
    this done?

26
Significant Findings (cont.)
  • Support maintenance models varies
  • All local, all remote w/ dispatch, or a blend
  • Backup, virus protection, security, DHCP, DNS, DC
  • Telco problem management is possible
  • Staff effort
  • Expect MAC calls at help desk about the same
  • Big Bonus 24 x 7 network monitoring problem
    resolution no additional cost
  • Ready access to additional technical depth

27
Significant Findings (cont.)
  • TDM vs. IP Telephony Replacement Cycle

Component
28
Significant Findings (cont.)
  • Uplift for student phones
  • Varied widely from College base (chart)
  • Per room per student also varied widely (chart)
  • Added 14 to project cost
  • 6 for VM, 8 for analog phone lines
  • Almost no increment to maintenance since opted to
    self-maintain gateways

29
Uplift for Analog Student Lines
30
Lessons Learned
  • Dont assume students no longer care about wired
    phones and that costs are prohibitive.
  • Real cost savings in convergence even for the
    smaller institution
  • Bidders have significantly different approaches
    to architecture, security, redundancy, Cat3
    cabling, and analog costs.
  • VARs with strong telephony and network expertise
    are more common there can be flexibility in
    crafting support agreements.

31
Contact Information
  • Mark J. Dumic
  • Associate Director Networking and
    Telecommunications
  • Swarthmore College
  • (610) 3288511
  • dumic_at_swarthmore.edu
Write a Comment
User Comments (0)
About PowerShow.com