Slide 1: members meeting

members meeting
8 feb 00


research tools and analyses




kc claffy, UCSD/SDSC/CAIDA
kc@caida.org 
www.caida.org

Slide 2: outline

outline


passive workload char.: coralReef

skitter
deployment status
asia connectivity study
dns study
other tools in library/package

hyperbolic graph/viz tool
current features
potential future features
solicit your input

Slide 3: workload characterization: coralReef

workload characterization: coralReef


insights for
usage profiling 
h/w, protocol, application design
architecture optimizing
capacity and peering planning 
network control and management 
security
performance analysis
delay, loss, jitter?
QOS assurance across ISPs
accounting and billing

tools: netramet, netflow, cflowd, coral
some suck less? ...evolution requires use

Slide 4: coralReef

coralReef

passive data analysis at 0C3,12,48,gigE
packet headers, filters for more
ATM, POS
integration w RRDtool, JChart (java), GD.pm for graphics
profile detection
SLA support

hardware side
dag3 (test, license)
dag4 (develop, test, license)
benchmark / calibrate accuracy

    www.caida.org/Tools/CoralReef

Slide 5: workload char of high perf. networks

workload char of high perf. networks

coral/ocXmon testing (oc3,12,48,gE)
persistent real-time full frame collection
integration w coralreef analysis s/w
dag4.0 testing `down'-underway

Slide 6: workload char: coralReef

workload char: coralReef

what to measure
how to viz
sample 19 aug 99, ucsd-cerfnet
https://anala.caida.org/CoralReef/Demos/

Slide 7

workload char.: protocol (proportion)

19 aug 99, ucsd-cerfnet

Slide 8: workload char: applications (ucsd-cerfnet)

workload char: applications (ucsd-cerfnet)

Slide 9: workload char.: mantra

workload char.: mantra
18 oct 99, fix-west.mbone.nasa.gov
http://www.caida.org/Tools/Mantra/

Slide 10: skitter: deployment status

skitter: deployment status


17 monitors
daily stats:     www.caida.org/Tools/Skitter/Summary/
make data sets available under AUP
root server project off to slow start
infrastructure
target lists harder
we're cycle-starved
initial asian study done
some correlation w routing data
need more disk space
need better destination gathering

Slide 11: skitter: preliminary findings

skitter: preliminary findings


~1% IP destinations disappearing monthly (re-addressing, firewalls)
route announced path not matching forward path
indication of potential routing configuration errors
    (by no means automatic)
persistence of paths
little correlation betw hop count & RTT
US fundamental for global transit 
methods to identify critical infrastructure
is there an Internet "core"?

datasets available to researchers

Slide 12: skitter: DNS root server study

skitter: DNS root server study


RSSAC, DNS technical advisory committee to ICANN
co-locate skitter hosts w root servers
demonstrate root server performance in serving target community
develop techniques for evaluating architectual optimality for root server placement 
great application for skitter
proactive participation
good cause

    www.caida.org/Tools/Skitter/RSSAC/

Slide 13: skitter: rtt distribution: tri-modal

skitter: rtt distribution: tri-modal

Slide 14: skitter: rtt vs longitude (light cone)

skitter: rtt vs longitude (light cone)

Slide 15: dispersion among ASes across paths

dispersion among ASes across paths

Slide 16: skitter: imminent tasks

skitter: imminent tasks


clean up destination lists
continue w metrics and analysis for `center', `core', `close'
     (optimize dist. server architectures)
metrics for `change'
better NetGeo support
revisit Asia study, compare 
continue root server study
correlation with routing, passive across time, sources, type
model of the Internet (no, really)

graphing/layout techniques....

Slide 17: hypview: 3D hyperbolic graph layout

hypview:  3D hyperbolic graph layout

potential future features
connect w external data srcs
point directly to cisco BGP feed
netflow
skitter
netgeo
routing registries
color, thickness, height of link/node by: 
country, #routes, traffic, delay, damage
drill-down 
click to pop up info about nodes, links
GUI to filter matching (unclear need)
`draw all paths thru this AS [&|to this prefix]'
`draw all paths > X [traffic|delay|routes]'
`draw all paths through ATM infrastructure'
`draw all paths not taking shortest paths'

Slide 18: hypview: as interface to the router

hypview: as interface to the router


histogram of AS paths through router
tabular summaries of 
routes vs traffic
AS path vs traffic
neighbors at levels of fanout
address space utilization

digestable routing table abstractions for
multicast (mbgp, msdp, whatever next)
animation of routing snapshots over time
correlated with traffic, performance, calendar
GUI to what-if interdomain simulations (!peer)  
need RIB

it's just another graphing tool...

Slide 19: overall: meas't & analysis challenges

overall:  meas't & analysis challenges


new methods for data collection, reduction, aggregation, mining, viz 
large, complex datasets ( ~Pbyte )
geographically and logically distributed 
dynamically changing
enable inter- and intra-ISP analysis 
       and feature detection
correlation among data sources/types
user-friendly integration with network utilities and control systems
proactive participation
top-down (app devel's scope constr.) 
bottom-up (ISP cooperation) 
vendors in middle (to right of research)

Slide 20: www.caida.org/Presentations/

www.caida.org/Presentations/

kc claffy
UCSD/SDSC/CAIDA
kc@caida.org
www.caida.org