Next Step In Signaling NSIS and Internet Routing Dynamics Charles Shen and Henning Schulzrinne Colum - PowerPoint PPT Presentation

1 / 3
About This Presentation
Title:

Next Step In Signaling NSIS and Internet Routing Dynamics Charles Shen and Henning Schulzrinne Colum

Description:

significant site to site variation - adaptive approach for NSIS ... Edge model: access routers of source and destination sites are NEs ... – PowerPoint PPT presentation

Number of Views:70
Avg rating:3.0/5.0
Slides: 4
Provided by: Nasir8
Category:

less

Transcript and Presenter's Notes

Title: Next Step In Signaling NSIS and Internet Routing Dynamics Charles Shen and Henning Schulzrinne Colum


1
Next Step In Signaling (NSIS) and Internet
Routing Dynamics Charles Shen and Henning
Schulzrinne _at_ Columbia University in the City of
New York
  • Project Description
  • As part of the IETF NSIS working group, we are
    standardizing a protocol for signaling
    information about a data flow along its path in
    the Internet.
  • The protocol supports various signaling
    applications, such as Quality of Service (QoS)
    and Network Address Translation (NAT) and
    Firewall traversal.
  • The protocol design adopts a two-layer approach,
    a lower layer for transport, and an upper layer
    specific to each signaling application.
  • We are looking specifically at the signaling
    transport layer, or Generic Internet Messaging
    Protocol for Signaling (GIMPS) as well as its
    interaction with IP routing.
  • We conducted Internet routing dynamics
    measurement and evaluated route change detection
    methods in typical NSIS deployment models.
  • Internet Routing Dynamics Analysis
  • Route prevalence and persistence
  • - single dominant route
  • - significant site to site variation
  • - adaptive approach for NSIS
  • Different types of route changes
  • - wide scales in time or location
  • - majority no change of hop count
  • - route splitting and load balancing
  • Measurement accuracy
  • - 10-min fixed vs. 2-hour exp.
  • - may capture the same changes
  • - may miss half of the changes.
  • - still site-to-site variation
  • Impact of multi-homing
  • - AS level route changes
  • - Asymmetric routing
  • End to end Routing Measurement
  • Traceroute for path characterization
  • 24 Public servers located in US, Iceland,
    Netherlands, Australia, Germany, Switzerland,
    Bulgaria, Sweden and Thailand.
  • Both independent, exponential interval (15/30
    min per site) and fixed interval (10 min per
    path) sampling
  • Between April and August 2004

NSIS-Concerned Route Changes - the subset of
route changes that involve change of NSIS
Entities (NEs)
New path
Data flow direction
Old path
  • NSIS Deployment Models and
  • the TTL Monitoring Method
  • AS model a central NE in each AS
  • Entry model ingress routers are NEs
  • Border model both ingress and egress routers
    are NEs
  • Edge model access routers of source and
    destination sites are NEs
  • TTL monitoring detected 40 - 90 route changes
    in different models

7DS Client
  • Route Change Detection Methods
  • Routing monitoring (BGP, OSPF)
  • - Local trigger
  • - Extended trigger
  • Packet monitoring (TTL/Interface)
  • - Connection-mode signaling messages
  • - Data packets
  • Datagram-mode probing

2
  • As part of the IETF NSIS working group we are
    standardizing a generic IP signaling protocol
    that separates the transport of the signaling
    from the application signaling (e.g., QoS, NAT
    traversal).

3
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com