/[gzz]/gzz/Documentation/misc/hemppah-progradu/research_problems
ViewVC logotype

Diff of /gzz/Documentation/misc/hemppah-progradu/research_problems

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.43 by hemppah, Wed Feb 5 12:51:44 2003 UTC revision 1.44 by hemppah, Wed Feb 5 13:39:24 2003 UTC
# Line 48  impossible Line 48  impossible
48  -Example applications: CFS \cite{dabek01widearea}, PAST \cite{rowstron01storage}, Oceanstore \cite{kubiatowicz00oceanstore}  -Example applications: CFS \cite{dabek01widearea}, PAST \cite{rowstron01storage}, Oceanstore \cite{kubiatowicz00oceanstore}
49  -data distribution approaches (like in Squirrel \cite{iyer02squirrel}): home node approach and directory approach  -data distribution approaches (like in Squirrel \cite{iyer02squirrel}): home node approach and directory approach
50  -CFS splits files into blocks (<50Kb), PAST distributed whole files  -CFS splits files into blocks (<50Kb), PAST distributed whole files
51    -the basic idea behind many DHTs is the fact that they perform operations in a binary-like tree
52    -more space/node --> the search arity is higher --> k is higher in k-ary trees
53    -DHTs' performance efficiency is derived from these tree based operations (e.g. split to half the previous scope)
54    
55    
56  *Update*  *Update*
57  -Viceroy system achieves O(log n) hops with only O(1) neighbors  -Viceroy system achieves O(log n) hops with only O(1) neighbors
# Line 159  Notice: pros and cons are not presented Line 163  Notice: pros and cons are not presented
163    
164  Summary  Summary
165    
166  This summary table is far from complete (and from total truth)  Following figures has to be done:
167    
168    1) application level overlay vs. network level overlay
169    2) Overview of structured overlay network
170    3) Overview of basic gnutella network
171    4) Overview of gnutella network with some improvements (super nodes and clusters)
172    5) Figure presenting k-ary tree (--> log n efficiency)
173    
174    Include following tables:
175    
176    1) Summary table:
177    
178  Protocol:               Insert/Delete   Space           Search                  # of network connections  Protocol:               Insert/Delete   Space           Search                  # of network connections
179  Chord:                  O(log^2 n)      O(log n)        O(log n)                2(log n)  Chord:                  O(log^2 n)      O(log n)        O(log n)                2(log n)
# Line 180  PeerNet 4):            O(log n)        O(log n)        O(log n) Line 194  PeerNet 4):            O(log n)        O(log n)        O(log n)
194  Kelips:                 *****           O(sqrt(n))      O(1)                    n/sqrt(n) + c*(sqrt(n)-1) + 'Total number of files'/sqrt(n)  Kelips:                 *****           O(sqrt(n))      O(1)                    n/sqrt(n) + c*(sqrt(n)-1) + 'Total number of files'/sqrt(n)
195  Freenet:                O(1)            O(1)            O(n)                    ??  Freenet:                O(1)            O(1)            O(n)                    ??
196    
197    
198    2) Difference table for approaches:
199    
200                                      DHT                           Broadcast
201    Queries:                          Controlled                    uncontrolled
202    A way for performing queries:     Exact keys                    Keywords
203    Query traffic:                    O(1)/O(log n)                 O(n)/O(n^2)
204    Will find the data item ?:        Yes                           Not necessarily
205    Overlay's structure:              Controlled and structured     Uncontrolled and ad hoc
206    Max. number of nodes ?:           Billions                      Millions
207    Data placement:                   Not local                     Local
208    Support for heterogeneiy:         No                            Yes
209    Support for locality:             Partial                       No
210    Possibilty for routing hotspots:  Yes                           No
211    Design/Implementation complexity: High                          Low
212    
213    
214    
215  * = In Kademlia, there is no action required when nodes leaves the system  * = In Kademlia, there is no action required when nodes leaves the system
216    
217  ** = Number of messages can grow as fast as O(n^2)  ** = Number of messages can grow as fast as O(n^2)

Legend:
Removed from v.1.43  
changed lines
  Added in v.1.44

savannah-hackers-public@gnu.org
ViewVC Help
Powered by ViewVC 1.1.26