Surveyor IP Performance Measurements

1
Surveyor IP Performance Measurements

• Matt Zekauskas
• matt_at_advanced.org
• June, 1999
• NLANR/I2 Joint Techs

2
IP Performance Metrics

• IETF IPPM effort
• Framework and Connectivity RFC
• One-way delay and packet loss drafts
• Others round-trip delay, bulk transfer, DV
• Surveyor implementation of one-way delay and
  packet loss metrics

3
Motivation for measuring delay

• Minimum of delay transmission/propagation delay
• Variation of delay queuing delay
• Large delay makes sustaining high-bandwidth flows
  harder
• Erratic variation in delay makes real-time apps
  harder

4
Uses

• Problem determination
• Engineering (trends, loads)
• Monitor QoS
• Feedback to advanced applications(for example,
  Tele-Immersion)

5
The Surveyor Infrastructure

• Measurement machines at campuses and at other
  interesting places along paths (e.g., gigaPoPs,
  interconnects)
• GPS to synchronize clocks
• Centralized database to store measurement data
• Web based reporting and analysis tools

6
Measurement Machines

• Dell 400 MHz Pentium Pro
• 128 MBytes RAM 8 GBytes disk
• BSDI Unix
• TrueTime GPS card and antenna (coax)
• Network Interface (10/100bT, FDDI, OC3 ATM in
  alpha)
• Special driver for the GPS card

7
Measurement Technology

• Active tests of one-way delay and loss
• Measurement daemon
• Test packets time-stamped with GPS time
• Back-to-back calibration 95 of measurements
  100 ?s ? 10 ?s soon
• Measurements centrally managed
• Concurrent routing measurements

8
Ongoing Tests - Routing

• Traceroute to same sites as One-Way delay
• Scheduled with Poisson process
• average rate one every 10 minutes

9
Central Database Machine

• Collects performance data from the measurement
  machines ssh, pull
• Stores the data in a home-grown database
• So far, all measurements taken on-line
• Serves data and summaries to reporting and
  analysis tools http

10
Current Surveyor Deployment

• 50 machines
• Universities
• Tele-Immersion Labs
• National Labs
• others

• 1883 paths
• NASA Ames XP
• I2 gigaPoPs (some)
• CAnet2 gigaPoPs
• APAN sites
• Abilene router nodes up with NTP, awaiting GPS

11
(No Transcript)
12
(No Transcript)
13
Reporting and analysis tools

• Web based tools
• Daily summary reports
• Integration with route measurements
• Java applet to dynamically graph almost ready
• Developing asynchronous notification of anomalies

14
Daily summary reports

• Take a 24-hour sample for a given path
• Divide it into one-minute sub-samples
• For each one-minute sub-sample
• Minimum delay (blue)
• 50th percentile (green)
• 90th percentile (red)

15
(No Transcript)
16
Lessons so far

• Routing is asymmetric
• Even when routing is symmetric,queuing is
  asymmetric
• Can detect level 2 changes (SONET failover or
  ATM)
• HPC connections do fall back to commodity paths,
  sometimes frequently or for long periods

17
Summary

• One-way Delay and Loss are
• practical
• useful
• Surveyor infrastructure growing
• Dynamic plotting, asynchronous notification
  coming soon
• Now focus on analysis and applications

18
More info

• Surveyor project info
• http//www.advanced.org/surveyor/
• Access to plots
• http//ippm-db.advanced.org/plots/
• IETF IPPM WG
• http//www.advanced.org/IPPM/
• My email matt_at_advanced.org