[chord] maintenance cost for Chord's finger tables.

[Hari Balakrishnan]

Your question is probably answered in
http://nms.lcs.mit.edu/papers/p2p-dynamic.html, in theory.

In practice, you can set the frequency to be whatever you want.  If it is
too low, the overlay may not be properly or fully connected.  If it is too
high, then you'd be wasting bandwidth.

The paper above shows that the right frequency depends on something called
the half-life of the P2P system.

-----Original Message-----
From: chord-bounces at amsterdam.lcs.mit.edu
[mailto:chord-bounces at amsterdam.lcs.mit.edu] On Behalf Of Hathai Tanta-ngai
Sent: Tuesday, October 12, 2004 3:04 PM
To: chord at amsterdam.lcs.mit.edu
Subject: [chord] maintenance cost for Chord's finger tables.


Hello there,

I'm working on service discoveries in P2P systems.
I found out that Chord has interesting properties to locate services. I
studied Chord from your paper entitled Chord: A Scalable Peer-to-Peer Lookup
Protocol for Internet Applications, appeared in IEEE/ACM TRANSACTIONS ON
NETWORKING. The paper does not mention about the maintenance cost of the
"stabilization" protocol which updates Chord's finger tables and sucessor
pointers. Could you plase give me the number that I can see the
state-management of Chord? I would like to analyze if I can extend the size
of the routing table and when the limitation occurs. Your feedback is really
appreciate.

Regards,
Hathai

###################################################
Hathai Tanta-ngai
Ph.D. Student, Faculty of Computer Science
Dalhousie University
Halifax, NS, Canada, B3H 1W5
Fax:   902-492-1517
Email: hathai at acm.org
#####################[THE END]#####################

_______________________________________________
chord mailing list
chord at amsterdam.lcs.mit.edu
https://amsterdam.lcs.mit.edu/mailman/listinfo/chord