CN103327046B - P2SP system scheduling methods, equipment and system based on node serve ability - Google Patents
P2SP system scheduling methods, equipment and system based on node serve ability Download PDFInfo
- Publication number
- CN103327046B CN103327046B CN201210076550.2A CN201210076550A CN103327046B CN 103327046 B CN103327046 B CN 103327046B CN 201210076550 A CN201210076550 A CN 201210076550A CN 103327046 B CN103327046 B CN 103327046B
- Authority
- CN
- China
- Prior art keywords
- node
- percent
- current
- service ability
- p2sp
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 34
- 238000011156 evaluation Methods 0.000 claims abstract description 46
- 238000013210 evaluation model Methods 0.000 claims description 20
- 238000013209 evaluation strategy Methods 0.000 claims description 2
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 8
- 230000005540 biological transmission Effects 0.000 description 7
- 238000012546 transfer Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 238000013467 fragmentation Methods 0.000 description 5
- 238000006062 fragmentation reaction Methods 0.000 description 5
- 238000007619 statistical method Methods 0.000 description 5
- 238000013461 design Methods 0.000 description 3
- 238000007596 consolidation process Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000013519 translation Methods 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
- 238000005303 weighing Methods 0.000 description 1
Abstract
Present invention is disclosed a kind of P2SP system scheduling methods, equipment and system based on node serve ability, the method comprising the steps of:Node serve capability evaluation strategic server receives the load information that each node reports in P2SP systems;The current service ability of the node is calculated according to the load information, and reports Tracker servers;Node scheduling is carried out according to the current service ability of the node for Tracker servers, and the result of node scheduling is fed back into the node.A kind of P2SP system scheduling methods, equipment and system based on node serve ability provided by the invention, improve the overall performance of P2SP networks.
Description
Technical field
The present invention relates to the communications field, more particularly to a kind of P2SP system call sides based on node serve ability
Method, equipment and system.
Background technology
P2SP (Peer to Server&Peer, put to server and point) technology is a kind of new network, relies on network
The computing capability and bandwidth of middle participant carries out file-sharing and download, rather than relies on several less servers.P2SP bases
In user to server and user's mechanism, different from P2P (Peer to Peer, point-to-point) technology, also different from P2S (Peer
To Server, put to server) technology, it not only supports P2P technologies, while also by searching database server resource
With P2P resource consolidations together, when user downloads a file, other resources are searched for automatically, select suitable resource to carry out
Accelerate, this causes P2SP in the stability of download and the speed downloaded, to there is very big raising than traditional P2P.
In P2SP technologies, the data source for downloading a file is divided into original link, P2P networks, third party mirror image
Auxiliary source, then file consolidation is together in series by the unique mark (such as MD5 or SHA) of complete file.Present P2SP
System architecture as shown in figure 1, download a file need following steps:
Node 95 downloads client or server and obtains URL (Uniform/ from internet or resource website
Universal Resource Locator, URL) link;
Node 95 inquires about multiple resource and file Hash using URL link as entrance, from Resource Server 91, is then looked into from Hash
The URL resource collection downloading datas ask;
Node 95 downloads completion post-registration fileinfo can pass through to Tracker servers 92, other nodes 95
Tracker services 92 inquire the node 95 downloaded the node 95 of completion and downloaded;
Other download nodes 95 start multi-source P2P and downloaded, and data are exchanged with each other between each node 95;
Download and the statistical information of correlation is reported into statistical server 93 after completing, complete P2SP downloading process.
As can be seen here, node 95 is used as entrance by a URL resource, with the help of background server, passes through resource
Server 91 can retrieve a collection of URL resources, and this batch of URL resource provides a variety of data for the download of client or server
The entrance in source, greatly improve the download performance of client or server and start the speed downloaded.
In existing P2SP systems, because high isomerism and dynamic, the upload band of each node 95 be present in each node 95
Width, concurrent connection number, the limitation of singular link bandwidth and download bandwidth limitation are all different.Tracker servers 92 are carrying out node
During scheduling, generally only consider operator belonging to node 95, the characteristic of Intranet and outer net carries out node scheduling, index is downloaded same
All nodes 95 of one file.When node 95 is inquired about, added in all current online nodes 95 according to node 95 suitable
Sequence, chosen node 95 and issue successively, when all nodes 95 be selected a wheel after, finish once wheel scheduling.Do not account for
The characteristic of network isomery, such as the node 95 of NAT (Network Address Translation, network address translation) network
Upper transfer performance does not have that transfer performance on the node of common network node 95 is good, the connectivity of 95 points than Intranet section of the connectivity of common network node 95
Good (Intranet node 95 usually requires secondary server and assists progress Firewall Traversing mutually to communicate).Thus to P2SP networks
Overall performance cause very big influence.
The content of the invention
The main object of the present invention is a kind of P2SP system scheduling methods based on node serve ability of offer, equipment and is
System, improve the overall performance of P2SP networks.
The present invention proposes a kind of P2SP system scheduling methods based on node serve ability, including step:
Node serve capability evaluation strategic server receives the load information that each node reports in P2SP systems;
The current service ability of the node is calculated according to the load information, and reports Tracker servers;For
Tracker servers carry out node scheduling according to the current service ability of the node, and the result of node scheduling is fed back into institute
State node.
Preferably, the service ability current according to load information calculate node is specially:
According to the current service ability of the load information and default node evaluation model calculate node.
Preferably, the node evaluation model is:
Rank=a*Uploadspeed+b*Downloadspeed+c*CurUpConnNum+d*CurD ownConnNum+
e*SigleConnBand;
Wherein, Rank is the current service ability of node, and Uploadspeed is the current average uploading speed of node,
Downspeed is the current average speed of download of node, and CurUpConnNum is that the current concurrently upload of node connects quantity,
CurDownConnNum is that the current parallel downloading of node connects quantity, and SigleConnBand is the current single link branch of node
The bandwidth held, a, b, c, d, e are respectively Uploadspeed, Downspeed, CurUpConnNum, CurDownConnNum
SigleConnBand weight.
Preferably, methods described also includes:
The statistical result that node serve capability evaluation strategic server is downloaded according to node, correct a, b, c, d, e institute
The weighted value of representative.
Preferably, the statistical result that the node is downloaded be to the speed of download of node, download time, download result, under
Load file size and/or Lifetime count obtained result.
The present invention also proposes a kind of P2SP system scheduling methods based on node serve ability, including step:
In the P2SP systems that the node serve capability evaluation strategic server that Tracker servers receive P2SP systems reports
The current service ability of each node;
Node scheduling is carried out according to the current service ability of the node.
Preferably, the service ability current according to node, which carries out node scheduling, includes:
Generate random value a Percent, 0 <=Percent <=1;
Each node is traveled through, as Percent ' >=Percent, the node belonging to Percent ' is chosen, and returns to life
A random value Percent is generated into unit, until the node of predetermined number is chosen;Wherein, Percent '=
SumPeerRank’/SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k is the nodes passed through
Amount, SumPeerRank=Rank1+Rank2+ ...+Rankn, n are total node number amount.
The present invention also proposes a kind of node serve capability evaluation strategic server based on P2SP systems, including:
Receiving module, for receiving the load information that each node reports in P2SP systems;
Computing module, for calculating the current service ability of the node according to the load information;
Reporting module, for reporting the Tracker servers of P2SP systems, for the Tracker servers according to
The current service ability of node carries out node scheduling, and the result of node scheduling is fed back into the node.
Preferably, the computing module is specifically used for:
According to the current service ability of the load information and default node evaluation model calculate node.
Preferably, the node evaluation model is:
Rank=a*Uploadspeed+b*Downloadspeed+c*CurUpConnNum+d*CurD ownConnNum+
e*SigleConnBand;
Wherein, Rank is the current service ability of node, and Uploadspeed is the current average uploading speed of node,
Downspeed is the current average speed of download of node, and CurUpConnNum is that the current concurrently upload of node connects quantity,
CurDownConnNum is that the current parallel downloading of node connects quantity, and SigleConnBand is the current single link branch of node
The bandwidth held, a, b, c, d, e be respectively Uploadspeed, Downspeed, CurUpConnNum,
CurDownConnNumSigleConnBand weight.
Preferably, the node serve capability evaluation strategic server also includes:
Correcting module, for the statistical result downloaded according to node, correct the weighted value representated by described a, b, c, d, e.
The present invention also proposes a kind of Tracker servers based on P2SP systems, including:
Receiving module, in the P2SP systems that the node serve capability evaluation strategic server for receiving P2SP systems reports
The current service ability of each node;
Scheduler module, for carrying out node scheduling according to the current service ability of the node.
Preferably, the scheduler module includes:
Generation unit, for generating random value a Percent, 0 <=Percent <=1;
Traversal Unit, for traveling through each node, as Percent ' >=Percent, the node quilt belonging to Percent '
Selection, and return to generation unit and generate a random value Percent, until the node of predetermined number is chosen;Wherein, Percent '
=SumPeerRank '/SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k is the nodes passed through
Amount, SumPeerRank=Rank1+Rank2+ ...+Rankn, n are total node number amount.
The present invention also proposes a kind of P2SP systems, including an at least Tracker servers, node serve capability evaluation strategy
Server and multiple nodes, wherein,
The node serve capability evaluation strategic server, the load information reported for receiving the node;And root
The current service ability of the node is calculated according to the load information, and reports Tracker servers;
The Tracker servers, for carrying out node scheduling according to the current service ability of the node, and by node
The result of scheduling feeds back the node.
Preferably, the node serve capability evaluation strategic server includes:
Receiving module, for receiving the load information that each node reports in P2SP systems;
Computing module, for calculating the current service ability of the node according to the load information;
Reporting module, for reporting the Tracker servers of P2SP systems, for the Tracker servers according to
The current service ability of node carries out node scheduling, and the result of node scheduling is fed back into the node.
Preferably, the computing module is specifically used for:
According to the current service ability of the load information and default node evaluation model calculate node.
Preferably, the node evaluation model is:
Rank=a*Uploadspeed+b*Downloadspeed+c*CurUpConnNum+d*CurD ownConnNum+
e*SigleConnBand;
Wherein, Rank is the current service ability of node, and Uploadspeed is the current average uploading speed of node,
Downspeed is the current average speed of download of node, and CurUpConnNum is that the current concurrently upload of node connects quantity,
CurDownConnNum is that the current parallel downloading of node connects quantity, and SigleConnBand is the current single link branch of node
The bandwidth held, a, b, c, d, e are respectively Uploadspeed, Downspeed, CurUpConnNum, CurDownConnNum
SigleConnBand weight.
Preferably, the node serve capability evaluation strategic server also includes:
Correcting module, for the statistical result downloaded according to node, correct the weighted value representated by described a, b, c, d, e.
Preferably, the Tracker servers include:
Receiving module, in the P2SP systems that the node serve capability evaluation strategic server for receiving P2SP systems reports
The current service ability of each node;
Scheduler module, for carrying out node scheduling according to the current service ability of the node.
Preferably, the scheduler module includes:
Generation unit, for generating random value a Percent, 0 <=Percent <=1;
Traversal Unit, for traveling through each node, as Percent ' >=Percent, the node quilt belonging to Percent '
Selection, and one random value Percent of generation is returned, until the node of predetermined number is chosen;Wherein, Percent '=
SumPeerRank’/SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k is the nodes passed through
Amount, SumPeerRank=Rank1+Rank2+ ...+Rankn, n are total node number amount.
A kind of P2SP system scheduling methods, equipment and system based on node serve ability proposed by the present invention, for existing
The problem of with the presence of technical scheme, node serve capability assessment model is introduced in P2SP download systems, because P2SP networks are worked as
In, the service ability of each node is different, for example uploading bandwidth, concurrently uploads connection number, be parallel downloading connection number, single-stranded
Tape splicing tolerance system, download bandwidth limitation etc., are all different.The present embodiment propose during node scheduling according to node
Real time load information carries out the model of node serve capability evaluation, and is taken in the P2SP download systems designed accordingly based on node
Business capacity of water carries out the tactful bandwidth resources integrity service ability that can make full use of node of node scheduling according to outline, reduces
Node start delay and data transfer delay, ensure the service quality that node is downloaded, lift the P2P nets of whole P2SP download systems
Network efficiency of transmission.
Brief description of the drawings
Fig. 1 is the system architecture diagram of the P2SP systems of prior art;
Fig. 2 is the schematic flow sheet of P2SP system scheduling methods one embodiment of the invention based on node serve ability;
Fig. 3 is the P2SP system scheduling methods one embodiment interior joint scheduling steps of the invention based on node serve ability
Schematic flow sheet;
Fig. 4 is the P2SP system scheduling methods one embodiment interior joint scheduling steps of the invention based on node serve ability
Schematic diagram;
Fig. 5 is the flow signal of the another embodiment of P2SP system scheduling methods one based on node serve ability of the present invention
Figure;
Fig. 6 is the structural representation of the embodiment of node serve capability evaluation strategic server one of the present invention;
Fig. 7 is the structural representation of Tracker servers in the embodiment of P2SP systems one of the present invention;
Fig. 8 is the structural representation of scheduler module in Tracker servers in the embodiment of P2SP systems one of the present invention;
Fig. 9 is the structural representation of the embodiment of P2SP systems one of the present invention.
The realization, functional characteristics and advantage of the object of the invention will be described further referring to the drawings in conjunction with the embodiments.
Embodiment
It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not intended to limit the present invention.
The problem of embodiment of the present invention exists for prior art, introduced in P2SP systems a kind of in node tune
The method that each node current service ability is assessed during degree, and devised according to this and node serve ability is based in P2SP systems
Size carries out the strategy of node scheduling according to outline.The strong node of service ability, the probability being selected are bigger.Pass through this scheduling plan
Slightly energy fully the bandwidth resources integrity service ability of sharp node, reduction node start delay and data transfer delay, ensures node
The service quality of download, lift the network transmission efficiency of whole P2SP download systems.
Reference picture 2, a kind of embodiment of P2SP system scheduling methods one based on node serve ability of the present invention is proposed, is used for
P2SP system interior joint service ability assessment strategy servers, including:
Step S10, the load information that each node reports in P2SP systems is received;
Each node downloads client or server periodically to node serve capability evaluation strategic server in P2SP systems
Online situation, the resource information locally possessed and current load information are reported, the load information includes uploading speed, downloads speed
Degree, upload and download the information such as concurrent connection number mesh.
Step S11, the current service ability of the node is calculated according to the load information;
Node serve capability evaluation strategic server is according to the current service ability of the load information calculate node, this reality
Example is applied preferably by the current service ability of a default node evaluation model evaluation node, and to Tracker server syncs
The result of node evaluation.
The node evaluation model of node serve ability can design as follows:
Assuming that node reports following load information:
Uploadspeed:Identify the average uploading speed of node;
Downspeed:Identify the average speed of download of node;
CurUpConnNum:The current concurrently upload linking number of node is identified, for example typically sets up 1 TCP channel as 1
Individual link;
CurDownConnNum:Identify the current parallel downloading linking number of node;
SigleConnBand:Node is identified currently when the bandwidth that a link is supported.
Then node evaluation model is Rank=F (a, b, c, d, e)=a*Uploadspeed+b*Downspeed+c*
CurUpConnNum +d*CurDownConnNum+e*SigleConnBand。
The probability of Rank ranking size of the principle of node selection based on node, each prioritizing selection belong to same ISP
The node high Rank of (Internet Service Provider, ISP), when actually realizing, Rank with
The situation of download that node reports and upload, can dynamic adjust, if such as a node concurrent connection number more than a fixed number
Mesh, in order to prevent overloading, its Rank can be reduced.The factor of above-mentioned node evaluation model can increase or delete according to being actually needed
Remove, to each decision Rank factor (Uploadspeed, Downspeed, CurUpConnNum, CurDownConnNum and
SigleConnBand) give a weighted value (a, b, c, d, e), among actual according to node scheduling result issue with node
The average speed of download of report constantly adjusts the weighted value of each factor, increases the weight of some factor or reduces another factor
Weight, it is final to determine one group of stable parameter.
Step S12, Tracker servers are reported;Entered for Tracker servers according to the current service ability of the node
Row node scheduling, and the result of node scheduling is fed back into the node.
When Tracker servers receive the download application of a registered node, Tracker servers are according to obtaining
The current service ability of each node in network, node scheduling is carried out, and by the IP address list of the node for possessing resource of selection
Feed back to the node filed an application.
Reference picture 3, node scheduling strategy are selection course of the Tracker servers to node, including:
Step S101, one random value Percent, Percent of generation is the probability that a node is selected.
Arbitrarily 2 random integers [M, N] of generation;
R '=M+ (int) ((N-M)+1) * RandX/ (Rand_Max+1);
P=R '-M;
Percent=(R '-M)/(N-M), 0=< percent <=1.
Step S102, enter the selection course that ingress selection section just represents a Peer every time, be used for weighing with Rank
One comprehensive value of node serve ability.As shown in figure 4, in Rank1 to Rankn, the line segment of different length represents different sections
The size of the ability of point, the service ability that line segment is longer to represent node is stronger, and line segment intersecting place is exactly Percent ', gives birth at random
Into Percent uniformly fall any point on line segment, the process of the selection of node is exactly to find Percent currently to fall at that
On individual line segment, Fig. 4 is intuitively drawn very much, line segment it is longer fall superincumbent possibility it is also bigger, i.e., the line segment represent node quilt
The probability of selection is also big.
Total service ability SumPeerRank=Rank1+Rank2 of all nodes is calculated when being selected
+ ...+Rankn, n represent the total number of node.SumPeerRank '=Rank1+Randk2 is calculated when a minor node selects
+ ... the interstitial content passed through during+Rankk, 1 <=k <=n, k expression selection node traverses, then calculate
Percent '=SumPeerRank '/SumPeerRank.Often calculated once by a node, occur Percent ' > first
When=Percent, the node for being identified as k is selected.
Step S103, judge to select whether the quantity of node reaches predetermined number, be then end node scheduling flow, otherwise
Return to step S101, carry out the selection of next node.
After the node that application is downloaded receives the node set of Tracker server feedbacks, to set interior nodes downloading data.
As found to download to the data fragmentation of completion in downloading process, then verified, data fragmentation mistake is such as found, to P2SP systems
Statistical server reported data error message as statistical information.After completing downloading task, this is reported to statistical server
What download time, speed of download, download result, file size, the original link downloaded and the different URL sources of downloading task obtained
The statistical information such as speed of download and download time, and as new URL sources, report Resource Server to enter resource database.
The download completion of statistical server recipient node and URL speeds of download, download time in downloading process, connect feelings
Condition, reporting for the statistical informations such as result, file size is downloaded, and the mode for being write as flowing water daily record supplies subsequent statistical analysis.Node
Service ability assessment strategy server is according to the result of statistical server statistical analysis, the related factor of amendment node evaluation model
And weight corresponding to each factor.
The problem of the present embodiment exists for prior art, introducing node serve ability is commented in P2SP download systems
Estimate model, among P2SP networks, the service ability of each node is different, for example uploading bandwidth, concurrently uploads connection
Number, parallel downloading connection number, the limitation of singular link bandwidth, download bandwidth limitation etc., are all different.The present embodiment propose
The model of node serve capability evaluation is carried out during node scheduling according to node real time load information, and design accordingly
Node can be made full use of by carrying out the strategy of node scheduling in P2SP download systems according to outline based on node serve capacity of water
Bandwidth resources integrity service ability, reduce node start delay and data transfer delay, ensure the service quality that node is downloaded, carry
Rise the P2P network transmission efficiencies of whole P2SP download systems.
Reference picture 5, a kind of embodiment of P2SP system scheduling methods one based on node serve ability of the present invention is proposed, is used for
Tracker servers in P2SP systems, including:
Step S50, respectively saved in the P2SP systems that the node serve capability evaluation strategic server of reception P2SP systems reports
The current service ability of point;
Each node downloads client or server periodically to node serve capability evaluation strategic server in P2SP systems
Online situation, the resource information locally possessed and current load information are reported, the load information includes uploading speed, downloads speed
Degree, upload and download the information such as concurrent connection number mesh.Node serve capability evaluation strategic server is according to default node evaluation
The current service ability of model calculate node, and to the result of Tracker server sync node evaluations.
Step S51, node scheduling is carried out according to the current service ability of the node.
Each node is current in the P2SP systems that Tracker server receiving node service ability assessment strategy servers report
Service ability Rank, when Tracker servers receive the download application of a registered node, Rank based on node row
The probability of name size, each prioritizing selection belong to the same ISP high nodes of Rank.Its selection course to node is as follows:
It is the probability that a node is selected to generate random value a Percent, Percent.
Arbitrarily 2 random integers [M, N] of generation;
R '=M+ (int) ((N-M)+1) * RandX/ (Rand_Max+1);
P=R '-M;
Percent=(R '-M)/(N-M), 0=< percent <=1.
As shown in figure 4, in Rank1 to Rankn, the line segment of different length represents the size of the ability of different nodes, line segment
It is longer represent node service ability it is stronger, line segment intersecting place is exactly Percent ', and the Percent generated at random uniformly falls
Any point on line segment, the process of the selection of node are exactly to find Percent currently to fall on that line segment, and Fig. 4 is very directly perceived
Ground draws, line segment it is longer fall superincumbent possibility it is also bigger, i.e., the selected probability of node that the line segment represents is also big.
Total service ability SumPeerRank=Rank1+Rank2 of all nodes is calculated when being selected
+ ...+Rankn, n represent the total number of node.SumPeerRank '=Rank1+Randk2 is calculated when a minor node selects
+ ... the interstitial content passed through during+Rankk, 1 <=k <=n, k expression selection node traverses, then calculate
Percent '=SumPeerRank '/SumPeerRank.Often calculated once by a node, occur Percent ' > first
When=Percent, the node for being identified as k is selected.
It is further continued for generating a random value Percent, the selection of next node is carried out, until the node of preset quantity is chosen
Select.After the node that application is downloaded receives the node set of Tracker server feedbacks, to set interior nodes downloading data.
The strategy for carrying out node scheduling according to outline based on node serve capacity of water that the present embodiment proposes, can be fully sharp
With the bandwidth resources integrity service ability of node, reduce node start delay and data transfer delay, ensure the clothes that node is downloaded
Business quality, lift the P2P network transmission efficiencies of whole P2SP download systems.
Reference picture 6, propose that a kind of node serve capability evaluation strategic server 100 1 based on P2SP systems of the present invention is real
Example is applied, including:
Receiving module 10, for receiving the load information that each node reports in P2SP systems;
Computing module 11, for calculating the current service ability of the node according to the load information;
Reporting module 12, for reporting the Tracker servers of P2SP systems, for the Tracker servers according to institute
State the current service ability of node and carry out node scheduling, and the result of node scheduling is fed back into the node.
Correcting module 13, for the statistical result downloaded according to node, correct described a, b, c, d, e.
Each node downloads client or server periodically to node serve capability evaluation strategic server in P2SP systems
100 receiving module 10 reports online situation, the resource information locally possessed and current load information, and the load information includes
The information such as uploading speed, speed of download, upload and download concurrent connection number mesh.Computing module 11 calculates according to above-mentioned load information
The current service ability of node, and to the result of Tracker server sync node evaluations, it is preferable that computing module 11 passes through
The current service ability of one default node evaluation model evaluation node, the node evaluation model of node serve ability can design
It is as follows:
Assuming that node reports following load information:
Uploadspeed:Identify the average uploading speed of node;
Downspeed:Identify the average speed of download of node;
CurUpConnNum:The current concurrently upload linking number of node is identified, for example typically sets up 1 TCP channel as 1
Individual link;
CurDownConnNum:Identify the current parallel downloading linking number of node;
SigleConnBand:Node is identified currently when the bandwidth that a link is supported.
Then node evaluation model is Rank=F (a, b, c, d, e)=a*Uploadspeed+b*Downspeed+c*
CurUpConnNum+d*CurDownConnNum+e*SigleConnBand。
Tracker servers belong to same ISP according to the probability of the Rank ranking sizes of each node, each prioritizing selection
The high nodes of Rank, when actually realizing, download and the situation of upload that Rank is reported with node, can dynamic adjust, such as
If a node concurrent connection number exceedes certain amount, in order to prevent overloading, its Rank can be reduced.Above-mentioned node evaluation mould
The factor of type can according to being actually needed increase or delete, to each decision Rank factor (Uploadspeed,
Downspeed, CurUpConnNum, CurDownConnNum and SigleConnBand) weighted value (a, b, c, d, e) is given,
The weighted value that the average speed of download reported with node constantly adjusts each factor is issued according to node scheduling result among reality,
Increase the weight of some factor or reduce the weight of another factor, it is final to determine one group of stable parameter.
When Tracker servers receive the download application of a registered node, Tracker servers are according to obtaining
The current service ability of each node in network, node scheduling is carried out, and by the IP address list of the node for possessing resource of selection
Feed back to the node filed an application.
After the node that application is downloaded receives the node set of Tracker server feedbacks, to set interior nodes downloading data.
As found to download to the data fragmentation of completion in downloading process, then verified, data fragmentation mistake is such as found, to P2SP systems
Statistical server reported data error message as statistical information.After completing downloading task, this is reported to statistical server
What download time, speed of download, download result, file size, the original link downloaded and the different URL sources of downloading task obtained
The statistical information such as speed of download and download time.
The download completion of statistical server recipient node and URL speeds of download, download time in downloading process, connect feelings
Condition, reporting for the statistical informations such as result, file size is downloaded, and the mode for being write as flowing water daily record supplies subsequent statistical analysis.Amendment
Module 13 is according to the result of statistical server statistical analysis, power corresponding to the related factor of amendment node evaluation model and each factor
Weight.
The problem of the present embodiment exists for prior art, introducing node serve ability is commented in P2SP download systems
Estimate model, among P2SP networks, the service ability of each node is different, for example uploading bandwidth, concurrently uploads connection
Number, parallel downloading connection number, the limitation of singular link bandwidth, download bandwidth limitation etc., are all different.The present embodiment propose
The model of node serve capability evaluation is carried out during node scheduling according to node real time load information, is available for Tracker to service
Device makes full use of the bandwidth resources integrity service ability of node based on node serve capacity of water, reduces node start delay sum
According to transmission delay, ensure the service quality that node is downloaded, lift the P2P network transmission efficiencies of whole P2SP download systems.
Reference picture 7, a kind of embodiment of Tracker servers 200 1 based on P2SP systems of the present invention is proposed, including:
Receiving module 21, the P2SP systems that the node serve capability evaluation strategic server for receiving P2SP systems reports
In the current service ability of each node;
Scheduler module 22, for carrying out node scheduling according to the current service ability of the node.
Reference picture 8, the scheduler module 22 include:
Generation unit 221, for generating random value a Percent, 0 <=Percent <=1;
Traversal Unit 222, for traveling through each node, as Percent ' >=Percent, the node belonging to Percent '
It is chosen, and returns to generation unit and generate a random value Percent, until the node of predetermined number is chosen;Wherein,
Percent '=SumPeerRank '/SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k is process
Number of nodes, SumPeerRank=Rank1+Rank2+ ...+Rankn, n are total node number amount.
Each node downloads client or server periodically to node serve capability evaluation strategic server in P2SP systems
Online situation, the resource information locally possessed and current load information are reported, the load information includes uploading speed, downloads speed
Degree, upload and download the information such as concurrent connection number mesh.Node serve capability evaluation strategic server is according to default node evaluation
The current service ability of model calculate node, and to the result of Tracker server sync node evaluations.
Each node is current in the P2SP systems that the receiving node service ability assessment strategy server of receiving module 21 reports
Service ability Rank, when Tracker servers receive the download application of a registered node, section is based on by scheduler module 22
The probability of the Rank ranking sizes of point, each prioritizing selection belong to the same ISP high nodes of Rank.
Scheduler module 22 is as follows to the selection course of node:
It is the probability that a node is selected that generation unit 221, which generates random value a Percent, Percent,.
Arbitrarily 2 random integers [M, N] of generation;
R '=M+ (int) ((N-M)+1) * RandX/ (Rand_Max+1);
P=R '-M;
Percent=(R '-M)/(N-M), 0=< percent <=1.
As shown in figure 4, in Rank1 to Rankn, the line segment of different length represents the size of the ability of different nodes, line segment
It is longer represent node service ability it is stronger, line segment intersecting place is exactly Percent ', and the Percent generated at random uniformly falls
Any point on line segment, the process of the selection of node are exactly to find Percent currently to fall on that line segment, and Fig. 4 is very directly perceived
Ground draws, line segment it is longer fall superincumbent possibility it is also bigger, i.e., the selected probability of node that the line segment represents is also big.
Total service ability SumPeerRank=Rank1+Rank2 of all nodes is calculated when being selected
+ ...+Rankn, n represent the total number of node.SumPeerRank '=Rank1+Randk2 is calculated when a minor node selects
+ ... the interstitial content passed through during+Rankk, 1 <=k <=n, k expression selection node traverses, then calculate
Percent '=SumPeerRank '/SumPeerRank.Often calculated once by a node, occur Percent ' > first
When=Percent, the node for being identified as k is selected.
Return to generation unit 221 and generate a random value Percent, the selection of next node is carried out, until preset quantity
Node be chosen.After the node that application is downloaded receives the node set of Tracker server feedbacks, downloaded to set interior nodes
Data.
The Tracker servers 200 that the present embodiment proposes are based on node serve capacity of water and carry out node tune according to outline
The strategy of degree, the bandwidth resources integrity service ability of node can be made full use of, reduce node start delay and data transfer delay,
Ensure the service quality that node is downloaded, lift the P2P network transmission efficiencies of whole P2SP download systems.
Reference picture 9, propose a kind of P2SP systems of the present invention, including an at least Tracker servers 200, node serve energy
Force estimation strategic server 100, statistical server 300, Resource Server 400 and multiple nodes 500, wherein, node serve
Capability evaluation strategic server 100 is used to receive the load information that each node 500 reports in P2SP systems;And according to preset
Node evaluation model calculates the current service ability of the node 500, and reports Tracker servers 200;Tracker is serviced
Device 200 is used to carry out node scheduling according to the current service ability of the node 500, and statistical server 300 is used for recipient node
The system such as URL speeds of download, download time, connection situation, download result, file size in 500 download completion and downloading process
Meter information reports, and the mode for being write as flowing water daily record supplies subsequent statistical analysis.Resource Server 400 is used for receiving node 500
The entrance for being used as index by URL inquires about file Hash and more URL resource collections;And file fragmentation check information is provided, supply
Download the validity of node verification downloading data.
The structure and operation principle of the present embodiment interior joint service ability assessment strategy server 100 are the same as the section shown in Fig. 6
Point service ability assessment strategy server 100, the structure and operation principle of Tracker servers 200 with shown in Fig. 7 or Fig. 8 it
Tracker servers 200, it is not repeated to describe herein.
It should be noted that in actual applications, node serve capability evaluation strategic server 100 can and statistical fractals
Any one is arranged on same server in device 300, Resource Server 400 or Tracker servers 200.
The preferred embodiments of the present invention are the foregoing is only, are not intended to limit the scope of the invention, every utilization
The equivalent structure or equivalent flow conversion that description of the invention and accompanying drawing content are made, or directly or indirectly it is used in other correlations
Technical field, be included within the scope of the present invention.
Claims (12)
1. a kind of P2SP system scheduling methods based on node serve ability, it is characterised in that including step:
Node serve capability evaluation strategic server receives the load information that each node reports in P2SP systems;
The current service ability of the node is calculated according to the load information, and reports Tracker servers;For Tracker
Server carries out node scheduling according to the current service ability of the node, and the result of node scheduling is fed back into the node;
Wherein, Tracker servers carry out node scheduling according to the current service ability of the node, including:
Generate random value a Percent, 0 <=Percent <=1;
Each node is traveled through, as Percent ' >=Percent, the node belonging to Percent ' is chosen, and it is single to return to generation
Member one random value Percent of generation, until the node of predetermined number is chosen;Wherein, Percent '=SumPeerRank '/
SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k are the number of nodes passed through, SumPeerRank
=Rank1+Rank2+ ...+Rankn, n are total node number amount;Node belonging to the Percent ' is corresponding to the Rankk
Node;
Ranki represents the current service ability of i-th of node, on average current according to i-th of node of the Ranki
Pass speed, current average speed of download, current concurrently upload connection quantity and current parallel downloading connection quantity meter
Obtain.
2. the method as described in claim 1, it is characterised in that the service ability current according to load information calculate node
Specially:
According to the current service ability of the load information and default node evaluation model calculate node.
3. method as claimed in claim 2, it is characterised in that the node evaluation model is:
Rank=a*Uploadspeed+b*Downloadspeed+c*CurUpConnNum+d*CurD ownConnNum+e*
SigleConnBand;
Wherein, Rank is the current service ability of node, and Uploadspeed is the current average uploading speed of node,
Downloadspeed is the current average speed of download of node, and CurUpConnNum is that the current concurrently upload of node connects number
Amount, CurDownConnNum are that the current parallel downloading of node connects quantity, and SigleConnBand is node currently single link
The bandwidth of support, a, b, c, d, e be respectively Uploadspeed, Downloadspeed, CurUpConnNum,
CurDownConnNumSigleConnBand weight.
4. method as claimed in claim 3, it is characterised in that also include:
The statistical result that node serve capability evaluation strategic server is downloaded according to node, correct representated by described a, b, c, d, e
Weighted value.
5. method as claimed in claim 4, it is characterised in that the statistical result that the node is downloaded is the download speed to node
Degree, download time, download the result that result, download file size and/or Lifetime count to obtain.
6. a kind of P2SP system scheduling methods based on node serve ability, it is characterised in that including step:
Respectively saved in the P2SP systems that the node serve capability evaluation strategic server that Tracker servers receive P2SP systems reports
The current service ability of point;
Node scheduling is carried out according to the current service ability of the node;
Wherein, the service ability current according to the node carries out node scheduling, including:
Generate random value a Percent, 0 <=Percent <=1;
Each node is traveled through, as Percent ' >=Percent, the node belonging to Percent ' is chosen, and it is single to return to generation
Member one random value Percent of generation, until the node of predetermined number is chosen;Wherein, Percent '=SumPeerRank '/
SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k are the number of nodes passed through, SumPeerRank
=Rank1+Rank2+ ...+Rankn, n are total node number amount;Node belonging to the Percent ' is corresponding to the Rankk
Node;
Ranki represents the current service ability of i-th of node, on average current according to i-th of node of the Ranki
Pass speed, current average speed of download, current concurrently upload connection quantity and current parallel downloading connection quantity meter
Obtain.
A kind of 7. node serve capability evaluation strategic server based on P2SP systems, it is characterised in that including:
Receiving module, for receiving the load information that each node reports in P2SP systems;
Computing module, for calculating the current service ability of the node according to the load information;
Reporting module, for reporting the Tracker servers of P2SP systems, for the Tracker servers according to the node
Current service ability carries out node scheduling, and the result of node scheduling is fed back into the node;
Wherein, Tracker servers carry out node scheduling according to the current service ability of the node, including:
Generate random value a Percent, 0 <=Percent <=1;
Each node is traveled through, as Percent ' >=Percent, the node belonging to Percent ' is chosen, and it is single to return to generation
Member one random value Percent of generation, until the node of predetermined number is chosen;Wherein, Percent '=SumPeerRank '/
SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k are the number of nodes passed through, SumPeerRank
=Rank1+Rank2+ ...+Rankn, n are total node number amount;Node belonging to the Percent ' is corresponding to the Rankk
Node;
Ranki represents the current service ability of i-th of node, on average current according to i-th of node of the Ranki
Pass speed, current average speed of download, current concurrently upload connection quantity and current parallel downloading connection quantity meter
Obtain.
8. node serve capability evaluation strategic server as claimed in claim 7, it is characterised in that the computing module is specific
For:
According to the current service ability of the load information and default node evaluation model calculate node.
9. node serve capability evaluation strategic server as claimed in claim 8, it is characterised in that the node evaluation model
For:
Rank=a*Uploadspeed+b*Downloadspeed+c*CurUpConnNum+d*CurD ownConnNum+e*
SigleConnBand;
Wherein, Rank is the current service ability of node, and Uploadspeed is the current average uploading speed of node,
Downloadspeed is the current average speed of download of node, and CurUpConnNum is that the current concurrently upload of node connects number
Amount, CurDownConnNum are that the current parallel downloading of node connects quantity, and SigleConnBand is node currently single link
The bandwidth of support, a, b, c, d, e be respectively Uploadspeed, Downloadspeed, CurUpConnNum,
CurDownConnNumSigleConnBand weight.
10. node serve capability evaluation strategic server as claimed in claim 9, it is characterised in that also include:
Correcting module, for the statistical result downloaded according to node, correct the weighted value representated by described a, b, c, d, e.
A kind of 11. Tracker servers based on P2SP systems, it is characterised in that including:
Receiving module, respectively save in the P2SP systems that the node serve capability evaluation strategic server for receiving P2SP systems reports
The current service ability of point;
Scheduler module, for carrying out node scheduling according to the current service ability of the node;
Wherein, scheduler module carries out node scheduling according to the current service ability of the node, including:
Generate random value a Percent, 0 <=Percent <=1;
Each node is traveled through, as Percent ' >=Percent, the node belonging to Percent ' is chosen, and it is single to return to generation
Member one random value Percent of generation, until the node of predetermined number is chosen;Wherein, Percent '=SumPeerRank '/
SumPeerRank;SumPeerRank '=Rank1+Rank2+ ...+Rankk, k are the number of nodes passed through, SumPeerRank
=Rank1+Rank2+ ...+Rankn, n are total node number amount;Node belonging to the Percent ' is corresponding to the Rankk
Node;
Ranki represents the current service ability of i-th of node, on average current according to i-th of node of the Ranki
Pass speed, current average speed of download, current concurrently upload connection quantity and current parallel downloading connection quantity meter
Obtain.
12. a kind of P2SP systems, it is characterised in that taken including an at least Tracker servers, node serve capability evaluation strategy
Business device and multiple nodes, wherein,
The node serve capability evaluation strategic server, the load information reported for receiving the node;And according to institute
State load information and calculate the current service ability of the node, and report Tracker servers;
The Tracker servers, for carrying out node scheduling according to the current service ability of the node, and by node scheduling
Result feed back the node;
The node serve capability evaluation strategic server is that the node serve ability any one of claim 7 to 10 is commented
Estimate strategic server, the Tracker servers are the Tracker servers described in claim 11.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210076550.2A CN103327046B (en) | 2012-03-21 | 2012-03-21 | P2SP system scheduling methods, equipment and system based on node serve ability |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210076550.2A CN103327046B (en) | 2012-03-21 | 2012-03-21 | P2SP system scheduling methods, equipment and system based on node serve ability |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103327046A CN103327046A (en) | 2013-09-25 |
CN103327046B true CN103327046B (en) | 2017-12-15 |
Family
ID=49195576
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210076550.2A Active CN103327046B (en) | 2012-03-21 | 2012-03-21 | P2SP system scheduling methods, equipment and system based on node serve ability |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103327046B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103957269B (en) * | 2014-05-08 | 2017-09-29 | 中国联合网络通信集团有限公司 | The system of selection of P2P network nodes and point-to-point redirection P2P Redirector servers |
CN105430028B (en) * | 2014-09-09 | 2019-03-29 | 阿里巴巴集团控股有限公司 | Service calling method, providing method and node |
CN104320272B (en) * | 2014-10-21 | 2019-03-15 | 中国联合网络通信集团有限公司 | A kind of method and the network equipment of transmission equipment information |
CN109962947B (en) * | 2017-12-22 | 2022-05-27 | 阿里巴巴集团控股有限公司 | Task allocation method and device in peer-to-peer network |
CN108683730B (en) * | 2018-05-11 | 2021-08-06 | 深圳市网心科技有限公司 | Resource scheduling method, service server, client terminal, network system and medium |
CN111064681B (en) * | 2018-10-16 | 2023-08-15 | 金山云(深圳)边缘计算科技有限公司 | Scheduling method and system |
CN111403013B (en) * | 2020-03-20 | 2024-02-02 | 西安翼展电子科技有限公司 | Method and device for capability assessment |
CN116095092B (en) * | 2023-04-11 | 2023-06-23 | 湖南快乐阳光互动娱乐传媒有限公司 | Peer node selection method and device |
Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101640699A (en) * | 2009-08-21 | 2010-02-03 | 深圳创维数字技术股份有限公司 | P2P streaming media system and streaming media downloading method thereof |
CN101957780A (en) * | 2010-08-17 | 2011-01-26 | 中国电子科技集团公司第二十八研究所 | Resource state information-based grid task scheduling processor and grid task scheduling processing method |
WO2011075291A1 (en) * | 2009-12-18 | 2011-06-23 | Alcatel-Lucent Usa Inc. | System and method for controlling peer-to-peer connections |
-
2012
- 2012-03-21 CN CN201210076550.2A patent/CN103327046B/en active Active
Patent Citations (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101640699A (en) * | 2009-08-21 | 2010-02-03 | 深圳创维数字技术股份有限公司 | P2P streaming media system and streaming media downloading method thereof |
WO2011075291A1 (en) * | 2009-12-18 | 2011-06-23 | Alcatel-Lucent Usa Inc. | System and method for controlling peer-to-peer connections |
CN101957780A (en) * | 2010-08-17 | 2011-01-26 | 中国电子科技集团公司第二十八研究所 | Resource state information-based grid task scheduling processor and grid task scheduling processing method |
Also Published As
Publication number | Publication date |
---|---|
CN103327046A (en) | 2013-09-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103327046B (en) | P2SP system scheduling methods, equipment and system based on node serve ability | |
JP6272451B2 (en) | Method and apparatus for implementing distributed content caching in a content distribution network | |
CN101247402B (en) | Multimedia files downloading and broadcasting system and method | |
EP3439239B1 (en) | Data transmission method and system based on peer-to-peer network | |
CN103248645A (en) | BT (Bit Torrent) off-line data downloading system and method | |
CN104052803A (en) | Decentralized distributed rendering method and system | |
CN102394899B (en) | On-demand system and method for improving file downloading speed | |
CN103634361B (en) | The method and apparatus for downloading file | |
CN103873507A (en) | Data block uploading and storing system and method | |
CN103957269A (en) | P2P network node selection method and P2P redirect server | |
CN101626389B (en) | Network node management method | |
CN103327121A (en) | P2P (Peer-to-Peer Network) network resource transmission method and device for same | |
CN103843297A (en) | Method, device and system for providing and selecting candidate nodes for live streaming services | |
CN105978936A (en) | CDN server and data caching method thereof | |
US20130110981A1 (en) | Peer-To-Peer Assist for Live Media Streaming | |
CN103297447A (en) | Resource sharing method and resource sharing equipment | |
CN106603660A (en) | Collaborative downloading method, terminal and system for distributed type videos | |
CN103812881A (en) | Offline downloading control method and apparatus | |
CN109962947A (en) | Method for allocating tasks and device in a kind of peer-to-peer network | |
CN114124981A (en) | Cloud storage data direct transmission management method, cloud transmission management platform and system | |
CN110839061B (en) | Data distribution method, device and storage medium | |
CN103916328B (en) | flow control method and system for content distribution network and peer-to-peer network | |
CN109474696B (en) | Network service method, device, electronic equipment and readable storage medium | |
CN102164115A (en) | Method and system for multisource mixing file transmission | |
CN113190583A (en) | Data acquisition system, method, electronic equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
TR01 | Transfer of patent right |
Effective date of registration: 20231228 Address after: 518000 Tencent Building, No. 1 High-tech Zone, Nanshan District, Shenzhen City, Guangdong Province, 35 Floors Patentee after: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. Patentee after: TENCENT CLOUD COMPUTING (BEIJING) Co.,Ltd. Address before: 2, 518044, East 403 room, SEG science and Technology Park, Zhenxing Road, Shenzhen, Guangdong, Futian District Patentee before: TENCENT TECHNOLOGY (SHENZHEN) Co.,Ltd. |
|
TR01 | Transfer of patent right |