Published on

Published in: Technology
1 Like
  • Be the first to comment

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide


  1. 1. IRM : I ntegrated File R eplication and Consistency M aintenance in P2P Systems A Technical Report Seminar Presented by K.SrikanthKumar 07F81A1243 Under the guidance of Mr. Ravikumar & co-ordinator Mr.b. ramakrishna
  2. 2. Abstract <ul><li>In peer to peer file sharing system, file replication and consistency maintenance are widely used techniques to improve the system performance. </li></ul><ul><li>Because of their interdependencies between them, these two issues are typically addressed separately. </li></ul><ul><li>This paper presents an Integrated file Replication and consistency Maintenance mechanism (IRM) that integrates the two techniques in a systematic and harmonized manner. </li></ul><ul><li>File replication methods specify replica nodes and consistency maintenance method propagate update messages based on message spreading. </li></ul><ul><li>Thus achieves high efficiency in file replication and consistency </li></ul><ul><li>In low cost. </li></ul>
  3. 3. introduction <ul><li>A recent large-scale characterization of HTTP traffic has shown that more than 75 percent of Internet traffic is generated by P2P applications. </li></ul><ul><li>The traffic in peer 2 peer has increased is increased due to wide availability of audio and video data. </li></ul>
  4. 4. What is file Replication? <ul><li>File replication is an effective method to deal with the problem of overload condition due to hot files. </li></ul><ul><li>File replication distributes loads over replica nodes to improve file query efficiency. </li></ul><ul><li>Replica Node: Replica node is that in which Replica files (copy of original files) exists. </li></ul><ul><li>In most current file replication methods, file owners rigidly specify replica nodes and the replica nodes passively accept replicas. </li></ul>
  5. 5. What is consistency maintenance? <ul><li>File consistency maintenance is to maintain consistency between files and its replicas. </li></ul><ul><li>The consistency Maintenance methods are is to update files. </li></ul><ul><li>The replica nodes may be continuously generated, deleted and fail, this leads to unsuccessful update propagation. </li></ul><ul><li>Thus file Replication should reduce to minimize consistency maintenance. </li></ul>
  6. 6. EXISTING APPROCH <ul><li>Due to interdependencies between file replication and consistency maintenance these two are addressed separately. </li></ul><ul><li>PROPOSED APPROCH </li></ul><ul><li>In proposed approach both file replication and consistency maintenance are integrated to improve the performance of the system in p2p </li></ul><ul><li>networks. </li></ul>
  7. 7. IRM : I ntegrated file R eplication and consistency M aintenance mechanisms . <ul><li>IRM: Mechanisms: </li></ul><ul><li>IRM file replication places replicas in frequently visited nodes to guarantee high utilization of replicas, and meanwhile reduce underutilized replicas. </li></ul><ul><li>Replica is created using query initiating rate and query passing rate. </li></ul><ul><li>IRM consistency maintenance in turn aims to guarantee file fidelity of consistency at a low cost with file replication dynamism consideration. </li></ul><ul><li>Using adaptive polling, IRM ensures timely update operations and avoids unnecessary updates. </li></ul>
  8. 8. IRM file replication and consistency maintenance diagram
  9. 9. Adaptive File Replication <ul><li>The replication algorithm achieves an optimized trade-off between query efficiency and overhead in file replication. </li></ul><ul><li>we introduce IRM’s file replication component by addressing two main problems in file replication: </li></ul><ul><li>1. Where to replicate files? </li></ul><ul><li>2. How to remove underutilized file replicas ? </li></ul>
  10. 10. <ul><li>Replica Nodes Determination: </li></ul><ul><li>Based on traffic load performance IRM replicates a file in nodes that have been very interested in the file or routing nodes that have been carrying more query traffic of the file. </li></ul><ul><li>Replica Creation: </li></ul><ul><li>We define a requester’s query initiating rate for file f, denoted by qf , as the number of queries for f sent by the requester during a unit time, say one second. </li></ul><ul><li>We define a node’s query passing rate of file f, denoted by lf , as the number of queries for file f received and forwarded by the node during a unit time. </li></ul><ul><li>Replica Adaptation: </li></ul><ul><li>IRM lets each replica node periodically update their query passing rate or query initiating rate of a file. If the rates are below their thresholds, the node removes the replica </li></ul>
  11. 11. File Consistency Maintenance <ul><li>IRM employs adaptive polling for file consistency maintenance to cater to file replication dynamism. </li></ul><ul><li>A poll approach puts the burden of consistency maintenance on individual nodes. </li></ul><ul><li>IRM addresses two main issues in consistency maintenance: </li></ul><ul><li>1. How to determine the frequency that a replica node probe a file owner in order to guarantee timely file update? </li></ul><ul><li>2. How to reduce the number of polling operations to save cost? </li></ul>
  12. 12. <ul><li>Polling Frequency Determination: </li></ul><ul><li>In case of file’s maximum update rate file replica can ensure that a replica is never outdated by more than t seconds by polling the owner every t seconds. </li></ul><ul><li>IRM associates a time-to-refresh (TTR) value with each replica . </li></ul><ul><li>The value is increased by an additive amount if the file doesn’t change between successive polls </li></ul><ul><li>TTR=TTR old +α </li></ul><ul><li>In the event the file is updated since the last poll, the TTR value is reduced by a multiplicative factor: </li></ul><ul><li>TTR=TTRold/ß </li></ul>
  13. 13. Poll Reduction: <ul><li>In addition to the file change rate, file query rate is also a main factor to consider in consistency maintenance. </li></ul>
  14. 14. Comparative Discussion of IRM <ul><li>File replication helps to minimize the number of replicas in order to minimize the overhead of consistency maintenance. </li></ul><ul><li>Consistency maintenance helps to guarantee the fidelity of consistency among replicas in file replication dynamism. </li></ul><ul><li>This harmonized integration helps IRM achieve high efficiency and effectiveness in both file replication and consistency maintenance . </li></ul>
  15. 15. The Impact of File Replication on Consistency Maintenance <ul><li>There are several traditional file replication methods that could generate more replicas due to variety of reasons </li></ul><ul><li>IRM minimizes the number of replicas while maintaining high efficiency and effectiveness of file replication </li></ul>
  16. 16. The Impact of File Replication on Consistency Maintenance <ul><li>In p2p their will be dynamism for nodes. </li></ul><ul><li>The structure may not be able to recover. </li></ul><ul><li>In IRM, a replica node pool the file server for update without the need of construct and maintain the structure </li></ul><ul><li>Unlike the previous consistency maintenance methods, which aim to update replicas soon after the original file is updated. </li></ul>
  17. 17. PERFORMANCE EVALUATION <ul><li>Experimental results show that IRM file replication algorithm is highly effective in reducing file query latency. </li></ul><ul><li>IRM file consistency maintenance provide guarantee of file fidelity. </li></ul><ul><li> Table: Simulated Environment and Algorithm Parameters </li></ul>
  18. 18. conclusion This paper proposes an IRM that achieves high efficiency at a significantly lower cost. Instead of passively accepting replicas and updates, nodes autonomously determine the need for file replication and validation based on file query rate and update rate. IRM reduces redundant file replicas, consistency maintenance overhead, and unnecessary file updates.
  19. 19. Future Enhancement <ul><li>We find that IRM relying on polling file owners still cannot guarantee that all file requesters receive up-to-date files, although its performance is better than other consistency maintenance algorithms. We plan to further study and explore adaptive polling methods to fully exploit file popularity and update rate for efficient and effective replica consistency maintenance </li></ul>