CN105357124B - A kind of MapReduce bandwidth optimization methods - Google Patents
A kind of MapReduce bandwidth optimization methods Download PDFInfo
- Publication number
- CN105357124B CN105357124B CN201510816378.3A CN201510816378A CN105357124B CN 105357124 B CN105357124 B CN 105357124B CN 201510816378 A CN201510816378 A CN 201510816378A CN 105357124 B CN105357124 B CN 105357124B
- Authority
- CN
- China
- Prior art keywords
- map
- openflow
- data packet
- flow table
- medians
- 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.)
- Expired - Fee Related
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/12—Shortest path evaluation
- H04L45/122—Shortest path evaluation by minimising distances, e.g. by selecting a route with minimum of number of hops
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/74—Address processing for routing
- H04L45/745—Address table lookup; Address filtering
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/50—Network services
- H04L67/60—Scheduling or organising the servicing of application requests, e.g. requests for application data transmissions using the analysis and optimisation of the required network resources
- H04L67/63—Routing a service request depending on the request content or context
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention discloses a kind of MapReduce bandwidth optimization methods, including:After Hadoop operations submission, the task execution nodal information that OpenFlow controllers are sent according to JobTracker, the node for executing Map medians and merging task is determined by Map median routing policies, the corresponding flow table item of update, and OpenFlow interchangers are sent to, OpenFlow interchangers are received and are installed to it by escape way, then carry out flow table item matching to the data packet received, if it is matched with Map median behavior types are merged, Map medians are merged.The present invention is by combining OpenFlow and Hadoop, utilize the data-handling capacity of OpenFlow interchangers, to Map medians, the intermediate tuple data obtained after the effect of map () function merges processing in advance, to alleviate the network congestion problem in data migration process significantly, Hadoop working efficiencies are significantly improved.
Description
Technical field
Present invention design belongs to Hadoop cloud calculating field, more particularly, to a kind of bandwidth optimization sides MapReduce
Method.
Background technology
MapReduce distribution calculating needs to carry out a large amount of one-to-many or multi-to-multi communication between server.This makes
In current technology environment, the situation of data center network congestion frequent occurrence increases so as to cause packet loss, propagation delay time increases
Big and throughput degradation.Especially Hadoop cloud calculate MapReduce during, when master server distribute Map tasks and
After Reduce tasks, Map servers proceed by evaluation work, and the median being calculated is moved to Reduce servers
It moves, a large amount of data are in transition process, it is easy to cause network congestion, keep Reduce Server latencies long or even lose
It loses.Data seriously reduce whole work in the time of network transmission and complete efficiency.For this problem, researcher proposes
Some prioritization schemes, such as MapReduce row storage optimizations, MapReduce connections optimize, MapReduce optimizing schedulings etc., but on
Scheme is stated to have the following problems mostly:Task scheduling is not separated by with resource allocation, scheduling deployment not enough automates, can not be fine
According to Network status Optimized Operation etc..
SDN, that is, Software Defined Network, i.e. software defined network, are existed by Stanford universities
Clean Slate are proposed in the works.Its design philosophy mainly separates network-based control function and forwarding capability, releases control
The coupling of plane processed and data plane so that control plane can more effectively provide unified network monitoring capability.SDN occurs
Later, academia begins attempt to realize distribution of the application layer to network flow, and most commonly used research is the cloud computing based on SDN
Data center's DCN network optimizations.Researcher proposes to combine SDN with Hadoop, utilizes the programmable network management of SDN
Advantage optimizes Hadoop network performances.Hadoop is broadly divided into following three aspects with the SDN prioritization schemes being combined at present:One
It is to improve Hadoop data transmission efficiencys by the way that queue priority is arranged, second is that the traffic aware and bandwidth using SDN divide in advance
With the characteristics of, third, pass through SDN improve Hadoop job scheduling algorithm.But these schemes can not be solved fundamentally
The limited problem of data transfer bandwidth during MapReduce.
The existing Hadoop prioritization schemes based on SDN are typically to be combined come reasonable distribution net with upper layer application by SDN
Network flow, or for the flow feature of Hadoop, the flow information of network layer is fed back into application layer, application layer passes through scheduling
Congestion path is evaded in the change of algorithm.But when a large amount of data flow is emerged in large numbers, evades congestion path and also powerlessly change congestion
Situation.
Invention content
For the disadvantages described above of the prior art, the present invention provides a kind of MapReduce bandwidth merged based on median
Optimization method combines SDN and Hadoop for the network congestion problem in MapReduce data migration process, utilizes
The data-handling capacity of OpenFlow interchangers, a part of Map medians of merging treatment in advance, i.e., after the effect of map () function
Obtained intermediate tuple data reduces the data traffic in shuffle stages in MapReduce operation process, so as to shorten data
Transit time improves Hadoop working efficiencies.
To achieve the above object, the present invention proposes a kind of MapReduce bandwidth optimization methods, which is characterized in that including
Following steps:
(1) after Hadoop operations submission, JobTracker sends task execution nodal information to OpenFlow controllers;
(2) OpenFlow controllers pass through the road of Value Data among Map according to the task execution nodal information of reception
The node for executing Map medians and merging task is further determined that by strategy, updates corresponding flow table item, which is issued to
OpenFlow interchangers;
(3) OpenFlow interchangers receive by escape way and install the flow table that OpenFlow controllers transmit
, and the data packet to receiving carries out flow table item matching, if data packet and flow table item successful match, execute and are instructed in flow table item
Behavior type merges Map medians.
As it is further preferred that in step (2), the Map medians routing policy is OpenFlow controller roots
It is obtained according to Hadoop job scheduling results.
As it is further preferred that in step (1), the task execution nodal information includes source address port, target
Address port, JobID, MapID, ReduceID.
As it is further preferred that the routing policy of Value Data is based on shortest-path rout ing algorithms among the Map
OpenFlow-MapReduce routing policies, specifically include following steps:
(2-1) determines source data packet in whole nodes of OpenFlow network topologies by shortest-path rout ing algorithms
The shortest path P of node s to destination node t0;
(2-2) judges P0In distributed map tasks number of nodes whether be more than 1, if it is not, going to step (2-3);If
It is then to further determine that P0In whether have not yet distribution Map medians merge task node, if so, be then followed successively by it is each
Each selected one of the nodes of Map tasks is distributed away from nearest neighbours and unappropriated node to appoint as executing Map medians and merge
The node of business, and network topology is updated, if nothing, go to step (2-3);
(2-3) excludes the path sought in OpenFlow network topologies, re-calls shortest path first searching
New shortest path P'0, judge P'0In distributed Map tasks number of nodes whether be more than 1, if so, selecting and exporting the road
Diameter is followed successively by each selected one of each node for having distributed Map tasks away from nearest neighbours and unappropriated node is as execution
Map medians merge the node of task, update network topology;If it is not, step (2-3) is then executed again, until all shortest paths
Path search finishes, if the number of nodes for not finding the task of distribution yet is more than 1 path, it is defeated to randomly select a shortest path
Go out, and updates network topology.
As it is further preferred that the step (3) specifically includes following steps:
(3-1) OpenFlow interchangers receive by escape way and install the flow table item that OpenFlow controllers transmit;
After (3-2) OpenFlow interchangers receive data packet, first data packet is added in caching, then it is carried out
Flow table item matches, if data packet matched arrive corresponding flow table item, jumps to step (3-3);If it is not, being then reported to OpenFlow controls
Device processed waits for working out and transfers to OpenFlow interchangers again after the routing policy of Value Data among Map;
(3-3) executes the instruction in flow table item, if it is forwarding to instruct behavior in flow table item, forwards data packet to corresponding mesh
Port;If it is existing behavior type in OpenFlow interchangers to instruct the behavior in behavior list, handed over according to OpenFlow
Intrinsic code of changing planes executes respective behavior;If it is not above two type to instruct behavior type, in decision instruction behavior list
Behavior type be type that system kernel can not be handled, jump to step (3-4);
(3-4) parses the MapReduce job informations of packet header, including JobID, MapID and ReduceID, and turns
Data packet format is changed, new data packets are obtained;
Whether (3-5) is directed to the new data packets in step (3-4), inquire with the presence of corresponding ReduceID queues, if depositing
The new data packets are being operated by Reduce then and are waiting for the merging of Map medians in insertion queue;If being not present,
It is medium pending that the new data packets are put into new queue;
Data packet after (3-6) merges Map medians carries out pipeline processes, matches flow table again, is transmitted to next
Jump OpenFlow interchangers.
As it is further preferred that existing behavior type includes regeneration behavior set, more in the OpenFlow interchangers
New data packets, update matching domain and more new metadata.
As it is further preferred that the new data packets include:JobID, former number belonging to former data packet, former data packet
The MapID of the ReduceID, former data packet that will be sent to according to packet, the Map numbers of merging, the MapID merged and data packet are deposited
Live time deadline.
In general, through the invention it is contemplated above technical scheme is compared with the prior art, mainly have below
Technological merit:
1. the present invention utilizes the data processing of OpenFlow interchangers by applying SDN in the MapReduce of Hadoop
Ability, a part of Map medians of merging treatment, reduce the data flow in shuffle stages in MapReduce operation process in advance
Amount, to substantially reduce the Data Migration time, significantly improves Hadoop working efficiencies;
2. the present invention by the Map median routing policies of proposition, further determines that executing Map medians merges task
Reduce nodes solve the network congestion problem in data migration process, are particularly suitable for mass data stream in network and emerge in large numbers
When application environment.
Description of the drawings
Fig. 1 is the MapReduce bandwidth optimization methods schematic diagram of the present invention under SDN and Hadoop system;
Fig. 2 is Map medians routing policy flow chart of the present invention;
Fig. 3 is the data packet matched flow table of the present invention and carries out Map median merging treatment flow charts;
Fig. 4 is Map medians merging treatment flow chart of the present invention.
Specific implementation mode
In order to make the purpose , technical scheme and advantage of the present invention be clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, and
It is not used in the restriction present invention.As long as in addition, technical characteristic involved in the various embodiments of the present invention described below
It does not constitute a conflict with each other and can be combined with each other.
Fig. 1 is the MapReduce bandwidth optimization methods schematic diagram of the present invention under SDN and Hadoop system.Including:
(1) after Hadoop operations submission, JobTracker sends task execution nodal information to OpenFlow controllers;
(2) OpenFlow controllers pass through the road of Value Data among Map according to the task execution nodal information of reception
The node for executing Map medians and merging task is further determined that by strategy, updates corresponding flow table item, which is issued to
OpenFlow interchangers;
(3) OpenFlow interchangers receive by escape way and install the flow table that OpenFlow controllers transmit
, and the data packet to receiving carries out flow table item matching, if data packet and flow table item successful match, execute and are instructed in flow table item
Behavior type merges Map medians.
Fig. 2 gives OpenFlow controllers during the present invention is implemented and carries out the Map medians used when routing policy calculating
The routing policy of routing policy flow chart, the centres Map Value Data is the OpenFlow- based on shortest-path rout ing algorithms
MapReduce routing policies determine that Value Data reaches Reduce among Map by the routing policy of Value Data among the Map
The shortest path of server, and it is away from nearest neighbours and not occupied to be followed successively by each selected one of each node for having distributed Map tasks
The node used merges the node of task as Map medians are executed.
It specifically includes:
(2-1) determines source data packet in whole nodes of OpenFlow network topologies by shortest-path rout ing algorithms
The shortest path P0 of node s to destination node t;
(2-2) judges whether the number of nodes that map tasks have been distributed in P0 is more than 1, if it is not, going to step (2-3);If
Be then further determine that whether have in P0 not yet distribution Map medians merge task node, if so, be then followed successively by it is each
Each selected one of the nodes of Map tasks is distributed away from nearest neighbours and unappropriated node to appoint as executing Map medians and merge
The node of business, and network topology is updated, if nothing, go to step (2-3);
(2-3) excludes the path sought in OpenFlow network topologies, re-calls shortest path first searching
New shortest path P'0, judge P'0In distributed Map tasks number of nodes whether be more than 1, if so, selecting and exporting the road
Diameter is followed successively by each selected one of each node for having distributed Map tasks away from nearest neighbours and unappropriated node is as execution
Map medians merge the node of task, update network topology;If it is not, step (2-3) is then executed again, until all shortest paths
Path search finishes, if the number of nodes for not finding the task of distribution yet is more than 1 path, it is defeated to randomly select a shortest path
Go out, and updates network topology.
After Fig. 3-4 indicates that data packet enters interchanger, Map median merging is carried out after being matched to flow table item according to packet header
Whole process, the process include:
(3-1) OpenFlow interchangers receive by escape way and install the flow table item that OpenFlow controllers transmit;
After (3-2) OpenFlow interchangers receive data packet, first data packet is added in caching, then it is carried out
Flow table item matches, if data packet matched arrive corresponding flow table item, jumps to step (3-3);If it is not, being then reported to OpenFlow controls
Device processed waits for working out and transfers to OpenFlow interchangers again after the routing policy of Value Data among Map;
(3-3) executes the instruction in flow table item, if it is forwarding to instruct behavior in flow table item, forwards data packet to corresponding mesh
Port;If the behavior in instruction behavior list is existing behavior type in OpenFlow interchangers, including regeneration behavior set,
Updated data package, update matching domain and more new metadata then execute respective behavior according to the intrinsic code of OpenFlow interchangers;If
It is not above two type to instruct behavior type, then the behavior type in decision instruction behavior list, which is system kernel, to be handled
Type, jump to step (3-4);
(3-4) parses the MapReduce job informations of packet header, including JobID, MapID and ReduceID, and turns
Data packet format is changed, new data packets are obtained, the new data packets include:JobID, former number belonging to former data packet, former data packet
The MapID of the ReduceID, former data packet that will be sent to according to packet, the Map numbers of merging, the MapID merged and data packet are deposited
Live time deadline;
Whether (3-5) is directed to the new data packets in step (3-4), inquire with the presence of corresponding ReduceID queues, if depositing
The new data packets are being operated by Reduce then and are waiting for the merging of Map medians in insertion queue;If being not present,
It is medium pending that the new data packets are put into new queue;
Data packet after (3-6) merges Map medians carries out pipeline processes, matches flow table again, is transmitted to next
Jump OpenFlow interchangers.
It should be noted last that the above specific implementation mode is merely illustrative of the technical solution of the present invention and unrestricted,
Although being described the invention in detail with reference to preferred embodiment, it will be understood by those of ordinary skill in the art that, it can be right
Technical scheme of the present invention is modified or replaced equivalently, without departing from the spirit of the technical scheme of the invention and range,
It is intended to be within the scope of the claims of the invention.
Claims (6)
1. a kind of MapReduce bandwidth optimization methods, which is characterized in that include the following steps:
(1) after Hadoop operations submission, JobTracker sends task execution nodal information to OpenFlow controllers;
(2) OpenFlow controllers pass through the routing plan of Value Data among Map according to the task execution nodal information of reception
It slightly further determines that and executes the node that Map medians merge task, update corresponding flow table item, which is issued to
OpenFlow interchangers;
(3) OpenFlow interchangers receive by escape way and install the flow table item that OpenFlow controllers transmit, and
Flow table item matching is carried out to the data packet received, if data packet and flow table item successful match, execute in flow table item and instruct behavior
Type merges Map medians;
The routing policy of Value Data is the OpenFlow-MapReduce routings based on shortest-path rout ing algorithms among the Map
Strategy specifically includes following steps:
(2-1) determines source data packet node s in whole nodes of OpenFlow network topologies, by shortest-path rout ing algorithms
To the shortest path P of destination node t0;
(2-2) judges P0In distributed map tasks number of nodes whether be more than 1, if it is not, going to step (2-3);If so, into
One step determines P0In whether there are not yet distribution Map medians to merge the node of task, each distributed Map if so, being then followed successively by
Section of each selected one of the node of task away from nearest neighbours and unappropriated node as execution Map medians merging task
Point, and network topology is updated, if nothing, go to step (2-3);
(2-3) excludes the path sought in OpenFlow network topologies, re-calls shortest path first and finds newly
Shortest path P '0, judge P '0In distributed the number of nodes of Map tasks and whether be more than 1, if so, select and export the path, according to
Each selected one of the secondary node each to have distributed Map tasks is away from nearest neighbours and unappropriated node is as in execution Map
Between be worth merging task node, update network topology;If it is not, step (2-3) is then executed again, until entire shortest paths searching
It finishes, if the number of nodes for not finding the task of distribution yet is more than 1 path, randomly selects a shortest path output, and more
New network topology.
2. according to the method described in claim 1, it is characterized in that, in step (2), the Map medians routing policy is
OpenFlow controllers are obtained according to Hadoop job scheduling results.
3. according to the method described in claim 1, it is characterized in that, in step (1), the task execution nodal information includes
Source address port, destination address ports, JobID, MapID, ReduceID.
4. according to the method described in claim 1, it is characterized in that, the step (3) specifically includes following steps:
(3-1) OpenFlow interchangers receive by escape way and install the flow table item that OpenFlow controllers transmit;
After (3-2) OpenFlow interchangers receive data packet, first data packet is added in caching, flow table then is carried out to it
Item matching jumps to step (3-3) if data packet matched arrive corresponding flow table item;If it is not, being then reported to OpenFlow controls
Device waits for working out and transfers to OpenFlow interchangers again after the routing policy of Value Data among Map;
(3-3) executes the instruction in flow table item, if it is forwarding to instruct behavior in flow table item, forwards data packet to corresponding destination
Mouthful;If it is existing behavior type in OpenFlow interchangers to instruct the behavior in behavior list, according to OpenFlow interchangers
Intrinsic code executes respective behavior;If it is not above two type, the row in decision instruction behavior list to instruct behavior type
It is the type that system kernel can not be handled for type, jumps to step (3-4);
(3-4) parses the MapReduce job informations of packet header, including JobID, MapID and ReduceID, and converts number
According to packet format, new data packets are obtained;
Whether (3-5) is directed to the new data packets in step (3-4), inquire with the presence of corresponding ReduceID queues, and if it exists, then
The new data packets are operated to be inserted into queue by Reduce and wait for the merging of Map medians;It is if being not present, this is new
It is medium pending that data packet is put into new queue;
Data packet after (3-6) merges Map medians carries out pipeline processes, matches flow table again, is transmitted to next-hop
OpenFlow interchangers.
5. according to the method described in claim 4, it is characterized in that, existing behavior type includes in the OpenFlow interchangers
Regeneration behavior set, updated data package, update matching domain and more new metadata.
6. according to the method described in claim 4, it is characterized in that, the new data packets include:Former data packet, former data packet
The MapID of ReduceID, former data packet that affiliated JobID, former data packet will be sent to, the Map numbers of merging, merged
MapID and data packet time-to-live deadline.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510816378.3A CN105357124B (en) | 2015-11-22 | 2015-11-22 | A kind of MapReduce bandwidth optimization methods |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201510816378.3A CN105357124B (en) | 2015-11-22 | 2015-11-22 | A kind of MapReduce bandwidth optimization methods |
Publications (2)
Publication Number | Publication Date |
---|---|
CN105357124A CN105357124A (en) | 2016-02-24 |
CN105357124B true CN105357124B (en) | 2018-08-17 |
Family
ID=55332997
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201510816378.3A Expired - Fee Related CN105357124B (en) | 2015-11-22 | 2015-11-22 | A kind of MapReduce bandwidth optimization methods |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN105357124B (en) |
Families Citing this family (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110233798B (en) | 2018-03-05 | 2021-02-26 | 华为技术有限公司 | Data processing method, device and system |
CN110502337B (en) * | 2019-07-12 | 2023-02-07 | 上海交通大学 | Optimization system for shuffling stage in Hadoop MapReduce |
CN110795220B (en) * | 2019-10-25 | 2022-04-22 | 北京浪潮数据技术有限公司 | Task merging method and device and computer readable storage medium |
JP2021144571A (en) * | 2020-03-13 | 2021-09-24 | 富士通株式会社 | Information processing device, transmission control method, and communication program |
CN111490795B (en) * | 2020-05-25 | 2021-09-24 | 南京大学 | Intermediate value length isomerism-oriented encoding MapReduce method |
WO2022007587A1 (en) * | 2020-07-08 | 2022-01-13 | 华为技术有限公司 | Switch and data processing system |
CN114844781B (en) * | 2022-05-20 | 2023-05-09 | 南京大学 | Method and system for optimizing Shuffle performance for encoding MapReduce under Rack architecture |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103428031A (en) * | 2013-08-05 | 2013-12-04 | 浙江大学 | Inter-domain link fast failure recovery method based on software defined network |
CN104243240A (en) * | 2014-09-23 | 2014-12-24 | 电子科技大学 | SDN (self-defending network) flow measuring method based on Open Flow |
CN104468353A (en) * | 2014-12-26 | 2015-03-25 | 深圳市新格林耐特通信技术有限公司 | SDN based data center network flow management method |
CN104852867A (en) * | 2014-02-19 | 2015-08-19 | 华为技术有限公司 | Data packet forwarding method, device and system |
CN105071947A (en) * | 2015-07-06 | 2015-11-18 | 山东超越数控电子有限公司 | Link optimization method for SDN network |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101764835B (en) * | 2008-12-25 | 2012-09-05 | 华为技术有限公司 | Task allocation method and device based on MapReduce programming framework |
KR101350333B1 (en) * | 2009-11-03 | 2014-01-10 | 한국전자통신연구원 | network emulation system and method for on-line server test processing heavy traffic |
CN103106249B (en) * | 2013-01-08 | 2016-04-20 | 华中科技大学 | A kind of parallel data processing system based on Cassandra |
CN103327128A (en) * | 2013-07-23 | 2013-09-25 | 百度在线网络技术(北京)有限公司 | Intermediate data transmission method and system for MapReduce |
US9367366B2 (en) * | 2014-03-27 | 2016-06-14 | Nec Corporation | System and methods for collaborative query processing for large scale data processing with software defined networking |
CN104102533B (en) * | 2014-06-17 | 2017-07-18 | 华中科技大学 | A kind of Hadoop dispatching methods and system based on bandwidth aware |
CN104391748A (en) * | 2014-11-21 | 2015-03-04 | 浪潮电子信息产业股份有限公司 | Mapreduce computation process optimization method |
CN104461748B (en) * | 2015-01-04 | 2017-06-09 | 电子科技大学 | A kind of optimal localization tasks dispatching method based on MapReduce |
CN106326308B (en) * | 2015-07-03 | 2019-06-11 | 华中科技大学 | Data de-duplication method and system in a kind of net based on SDN |
-
2015
- 2015-11-22 CN CN201510816378.3A patent/CN105357124B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103428031A (en) * | 2013-08-05 | 2013-12-04 | 浙江大学 | Inter-domain link fast failure recovery method based on software defined network |
CN104852867A (en) * | 2014-02-19 | 2015-08-19 | 华为技术有限公司 | Data packet forwarding method, device and system |
CN104243240A (en) * | 2014-09-23 | 2014-12-24 | 电子科技大学 | SDN (self-defending network) flow measuring method based on Open Flow |
CN104468353A (en) * | 2014-12-26 | 2015-03-25 | 深圳市新格林耐特通信技术有限公司 | SDN based data center network flow management method |
CN105071947A (en) * | 2015-07-06 | 2015-11-18 | 山东超越数控电子有限公司 | Link optimization method for SDN network |
Also Published As
Publication number | Publication date |
---|---|
CN105357124A (en) | 2016-02-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105357124B (en) | A kind of MapReduce bandwidth optimization methods | |
US10263889B2 (en) | Data forwarding method, device, and system in software-defined networking | |
CA2882535C (en) | Control device discovery in networks having separate control and forwarding devices | |
US9369350B2 (en) | Method and system of network transfer adaptive optimization in large-scale parallel computing system | |
CN108260169B (en) | QoS guarantee-based dynamic service function chain deployment method | |
CN107094115B (en) | Ant colony optimization load balancing routing algorithm based on SDN | |
CN100596102C (en) | Method for establishing label switched path of minimized path preemption cost | |
CN106789648B (en) | Software defined network route decision method based on content storage and Network status | |
US9367366B2 (en) | System and methods for collaborative query processing for large scale data processing with software defined networking | |
JP2022532729A (en) | Slice-based routing | |
Wang et al. | Freeway: Adaptively isolating the elephant and mice flows on different transmission paths | |
CN107294852B (en) | Network routing method using topology dispersed short path set | |
CN108134967B (en) | Power optical network routing frequency spectrum distribution method and system | |
CN106209669A (en) | Towards SDN data center network maximum of probability path stream scheduling method and device | |
CN106936705B (en) | Software defined network routing method | |
CN109614215A (en) | Stream scheduling method, device, equipment and medium based on deeply study | |
CN105553845A (en) | Software defined network route selection method | |
CN104754589B (en) | Dynamic networking method based on traffic aware | |
KR20150080183A (en) | Method and Apparatus for dynamic traffic engineering in Data Center Network | |
Jiang et al. | Tailor: Trimming coflow completion times in datacenter networks | |
JP2006287919A (en) | Communication network, content distribution node, tree construction method, and content distribution control program | |
JP2012175480A (en) | Communication control device, communication control method and communication control program | |
CN114938327A (en) | Routing method, routing device, controller and computer readable storage medium | |
CN105610712B (en) | The method for reducing whole network data stream forward delay based on software defined network framework | |
CN114448879B (en) | SDN-based data center network flow scheduling method |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20180817 Termination date: 20191122 |
|
CF01 | Termination of patent right due to non-payment of annual fee |