1. 1 A Large-Scale Network Testbed Jay Lepreau Chris Alfeld David Andersen Kevin Van Maren University of Utah http://www.cs.utah.edu/flux/testbed/ September 2, 1999

2. 2 Prototype Pieces: 105 edge nodes

3. 3 Experimental platforms for network and distributed systems research: Today, 3 options: simulation, small static testbeds, live networks. Each has well known limitations… Hopefully someday, a large scale set of distributed small testbeds, attracting users (“Access”) Soon, a new option: a medium-to-large-scale, entirely configurable testbed This is a “network for research” Also presents challenging research issues of its own

4. 4 Some Unique Characteristics Significant scale: initially 225 nodes, degree four 100Mb links between 42 core routers. User-configurable control of “physical” characteristics: shaping of link latency/bandwidth/drops/errors (via invisibly interposed “shaping nodes”), router processing power, buffer space, … Node breakdown: 42 core, 160 edge, 21 shaping, 2 management

5. 5 A View from the Dark Side

6. 6 And the Light Side

7. 7 More Unique Characteristics Capture of low-level node behavior such as interrupt load and memory bandwidth User-replaceable node OS software User-configurable physical link topology (VLAN via BFS; “P-LAN” via BFPP) Completely configurable and usable by external researchers, including node power cycling

8. 8 Roatan: Remote Console for a Node

9. 9 Feature: Automatic mapping of desired topologies and characteristics to physical resources Algorithm goals: Üminimize likelihood of experimental artifacts (bottlenecks) Ü“optimal” packing of multiple simultaneous experiments Ü Complete in finite time! Constraint-based heuristic algorithm (version 2!) Feature: accepts ns-compatible specification

10. 10 Early Network Configuration GUI

11. 11 Research Applications Simulation validation Active networks Resource demands of services inside routers Denial-of-service resistance Interaction of adaptive applications and protocols All sorts of distributed system experiments...

12. 12 Research Issues and Other Challenges Calibration, validation, and scaling: how to emulate different speed networks? Scaling behavior of emulating faster links by slowing nodes? Can we sufficiently capture real router internal behavior in a PC? Assuring validity: detecting switch bottlenecks, measuring and controlling physical characteristics without introducing artifacts. Algorithms and software to map requirements to resources while minimizing artifacts. Integrate with ns? Providing a reasonable user interface to all this.

13. 13 Status Hardware and software design, remote management and GUI, mapping software vers. 1, starting to build racks ~16000 lines of code so far (Tcl, C++) Most of budget in place for a prototype version

14. 14 Open to and Looking for... Feedback Ü Examples where current methods fail and this would help Ü Examples of how this one loses! And other advice. Early users Collaborators on both the R and D issues of testbed Shameless plugs: ÜEquipment donations (BFSwitch, esp. slightly programmable, PCs, NICs), $s ÜUtah hiring networking faculty