/[fenfire]/journals/plans/SA_2003.rst
ViewVC logotype

Diff of /journals/plans/SA_2003.rst

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

revision 1.128 by jvk, Wed May 14 10:41:19 2003 UTC revision 1.129 by hemppah, Wed May 14 10:44:50 2003 UTC
# Line 589  basic research. Line 589  basic research.
589  Efficient information sharing in peer-to-peer environment  Efficient information sharing in peer-to-peer environment
590  '''''''''''''''''''''''''''''''''''''''''''''''''''''''''        '''''''''''''''''''''''''''''''''''''''''''''''''''''''''      
591    
592  Efficient resource discovery is an essential problem in peer-to-peer networks because there is  Efficient resource discovery is an essential problem in Peer-to-Peer networks because there is
593  no centralized index where to look for information about resources. There are two well known  no centralized index where to look for information about resources. There are two well known
594  approaches in which all modern Peer-to-Peer systems fall: the loosely structured approach  approaches in which all modern Peer-to-Peer systems fall: the loosely structured approach
595  and the tightly structured approach. In the loosely structured approach the construction  and the tightly structured approach [hyytiala03stormp2p]_. In the loosely structured approach the
596  and the maintenance of the network is controlled loosely while tightly structured approach  construction and the maintenance of the network is controlled loosely. In the loosely structured approach,
597    t he performance of resource discovery is $O(n)$ and the number of network packets can grow with
598    $O(n^{2})$, where $n$ is the number of participating nodes in the network. Tightly structured approach
599  is constructed deterministically, which all participating peers have to follow; the topology of the  is constructed deterministically, which all participating peers have to follow; the topology of the
600  overlay and the placement of services is controlled tightly. Currently either of the approches  overlay and the placement of services is controlled tightly. This feafure provides $O(\log{n})$
601  do not support *both* adaptable and efficient resource discovery in peer-to-peer environment.  resource discovery efficiency and the number of packets grow with $O(\log{n})$. However,
602    the tightly structured approach may not be reliable in highly adverse conditions, such as
603  In our research, we    sudded network partition or wide hostile attack. Thus, either of the approches do not support
604    *both* adaptable and efficient resource discovery in peer-to-peer environment.
605    
606  Currently we use distributed hash tables for searching blocks  In our research, we focus on developing algorithms which are both adaptable and efficient in highly
607  in a distributed environment, but the use of other Peer-to-Peer technologies are  adverse conditions. Specifically, we study if we can combine the adaptivity of the loosely
608  also studied.  structured approach with the efficiently of the tightly structured approach. We also examine
609    how well distributed resource discovery algorithms can be made to endurate hostile attacks
610  Research on how location-independent identifiers can be used to find  in Peer-to-Peer enviroment.
 information in a Peer-to-Peer environment. To support our research work, we have  
 developed a software library, named as Storage Modules (Storm), which provides  
 the basis for storing (xanalogical) data to our system. Storm provides an alternative to  
 traditional file-based storage as all data is stored as immutable *blocks*. A block is  
 like a file but it cannot be changed, and instead of a file name it is identified by a  
 SHA-1 cryptographic content hash [fips-sha-1]_. Thus, identifiers are  
 location-independent and globally unique.  
         
 Research focuses on finding efficient methods for locating xanalogical media  
 using location-independent and globally unique identifiers from the Peer-to-Peer  
 network. Currently we use distributed hash tables for searching blocks  
 in a distributed environment, but the use of other Peer-to-Peer technologies are  
 also studied.  
   
 .. more about theories behind the p2p stuff (HH)  
     
 Our goal is to build a Peer-to-Peer system that allows one to perform a *global*  
 data lookups and fetch a block if *anybody* has published it in the Internet.  
   
   
   
 Flexible focus+context views such as fisheye distortions  
 allow more context to be shown but also induce  
 usability issues that need to be solved.  
611    
612        
613  Research schedule  Research schedule

Legend:
Removed from v.1.128  
changed lines
  Added in v.1.129

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