CN103678235B - Based on parallel pipeline network data processing device and method - Google Patents
Based on parallel pipeline network data processing device and method Download PDFInfo
- Publication number
- CN103678235B CN103678235B CN201310662519.1A CN201310662519A CN103678235B CN 103678235 B CN103678235 B CN 103678235B CN 201310662519 A CN201310662519 A CN 201310662519A CN 103678235 B CN103678235 B CN 103678235B
- Authority
- CN
- China
- Prior art keywords
- data
- module
- network
- network interface
- feedback
- 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
Abstract
The present invention relates to a kind of based on parallel pipeline network data processing device and method, does it comprise? PCIE? X8 EBI, described PCIE? X8 EBI is connected with arbitration modules by cache module, described arbitration modules is connected with transceiver by storage device, described transceiver is connected with network interface transceiving data device, described storage device comprises at least two memory modules, in transceiver, comprise and in storage device, store the corresponding consistent sending/receiving module of module number, in network interface transceiving data device, comprise and the corresponding consistent network interface transceiving data module of memory module quantity, sending/receiving module in memory module and transceiver in storage device connects one to one, and memory module connects one to one by sending/receiving module and network interface transceiving data module. compact conformation of the present invention, can realize the network data processing of parallel pipeline, improves the speed of data processing, and wide accommodation is safe and reliable.
Description
Technical field
The present invention relates to a kind of data processing equipment and method, especially a kind of based on parallel pipeline network data processing device and method, belong to the technical field of network data processing.
Background technology
Parallel transmission refers to that long numeric data transmits by parallel line simultaneously, can greatly improve the speed of transfer of data. Streamline refers to that in the time that program is carried out the accurate parallel processing of one that many effects of overlappings operate realizes technology. How in the time of network data, realizing parallel pipelining process line technology is a current difficult problem.
Summary of the invention
The object of the invention is to overcome the deficiencies in the prior art, provide a kind of based on parallel pipeline network data processing device and method, its compact conformation, can realize the network data processing of parallel pipeline, improve the speed of data processing, wide accommodation, safe and reliable.
According to technical scheme provided by the invention, described based on parallel pipeline network data processing device, comprise PCIEX8 EBI, described PCIEX8 EBI is connected with arbitration modules by cache module, described arbitration modules is connected with transceiver by storage device, described transceiver is connected with network interface transceiving data device, described storage device comprises at least two memory modules, in transceiver, comprise and in storage device, store the corresponding consistent sending/receiving module of module number, in network interface transceiving data device, comprise and the corresponding consistent network interface transceiving data module of memory module quantity, sending/receiving module in memory module and transceiver in storage device connects one to one, and memory module connects one to one by sending/receiving module and network interface transceiving data module,
PCIEX8 EBI deposits the transmission data that receive in cache module in, in the time that arbitration modules detects that cache module has the data of transmission, the data head that sends data in cache module is resolved, after transmission data all deposit in cache module, arbitration modules selects corresponding network interface transceiving data module to carry out data transmission according to the data head that sends data; In the time that network interface transceiving data module detects network feedback data, packet number and the byte number of sending/receiving module statistics network feedback data, and network feedback data is all deposited in the memory module being connected with described sending/receiving module together with packet number and the byte number of statistics; Arbitration modules reads the data of storage in memory module and adds required data head and receives data to form network-feedback, and described network-feedback reception data are deposited in cache module, reads for PCIEX8 EBI.
Described arbitration modules is numbered the network interface transceiving data module in network interface transceiving data device, and the data head that arbitration modules is added comprises the numbering that is connected network interface transceiving data module with memory module, packet number and the byte number of network feedback data.
Described PCIEX8 EBI is connected with host computer, and the network-feedback that host computer reads in cache module by PCIEX8 EBI receives data, and can transmit required transmission data to PCIEX8 EBI.
A kind of based on parallel pipeline network data processing method, described network data processing method comprises the steps:
A, receive and send data by PCIEX8 EBI, and described transmission data are deposited in cache module;
B, arbitration modules detect in real time to cache module, and in the time having data in cache module, arbitration modules is resolved the data head that sends data in cache module; Arbitration modules judges according to the data head that sends data whether completely enter cache module, after transmission data enter in cache module completely, arbitration modules deposits transmission data in corresponding memory module according to the data head that sends data if sending data;
C, the sending/receiving module being connected with memory module detect and send after data, and sending/receiving module will send data and will send data by network interface transceiving data module and outwards send;
D, by network interface transceiving data module receive network feedback data, sending/receiving module is added up packet number, the byte number of network feedback data, and by network feedback data together with statistics packet number, byte number all deposit in memory module;
E, arbitration modules read network feedback data, packet number, the byte number in memory module, and according to the numbering of the network interface transceiving data module being connected with memory module, network feedback data are added to data head, receive data to form network-feedback;
F, arbitration modules receive data by network-feedback and deposit in cache module, read described network-feedback receive data for PCIEX8 EBI.
When having while all having network feedback data in multiple memory modules, arbitration modules reads successively according to the order of network interface transceiving data module numbering.
In described step b, arbitration modules is resolved the numbering of the network interface transceiving data module that obtains sending packet number, the byte number of data and outwards send to sending the data head of data.
Advantage of the present invention: receive transmission data by PCIEX8 EBI, arbitration modules receives sending data, judgement and parsing, and send by corresponding network interface transceiving data module, the network feedback data that arbitration modules receives network interface transceiving data module is added data head and is formed network-feedback reception data, network-feedback is received to data to be deposited in cache module, after reading for PCIEX8 EBI, be transferred to host computer, its compact conformation, can realize the network data processing of parallel pipeline, improve the speed of data processing, wide accommodation, safe and reliable.
Brief description of the drawings
Fig. 1 is structured flowchart of the present invention.
Fig. 2 is the flow chart that arbitration modules of the present invention deposits transmission data in storage device.
Fig. 3 is the flow chart that the present invention's the first network interface transceiving data module is carried out data transmission.
Fig. 4 is that arbitration modules of the present invention is to receiving the flow chart of data processing.
Description of reference numerals: 1-PCIEX8 EBI, 2-cache module, 3-arbitration modules, 4-storage device, 5-transceiver and 6-network interface transceiving data device.
Detailed description of the invention
Below in conjunction with concrete drawings and Examples, the invention will be further described.
As shown in Figure 1: in order to realize by a PCIEX8 EBI 1 network data processing of parallel pipeline, the present invention includes PCIEX8 EBI 1, described PCIEX8 EBI 1 is connected with arbitration modules 3 by cache module 2, described arbitration modules 3 is connected with transceiver 5 by storage device 4, described transceiver 5 is connected with network interface transceiving data device 6, described storage device 4 comprises at least two memory modules, in transceiver 5, comprise and the corresponding consistent sending/receiving module of the interior storage module number of storage device 4, in network interface transceiving data device 6, comprise and the corresponding consistent network interface transceiving data module of memory module quantity, sending/receiving module in memory module in storage device 4 and transceiver 5 connects one to one, and memory module connects one to one by sending/receiving module and network interface transceiving data module,
PCIEX8 EBI 1 deposits the transmission data that receive in cache module 2 in, in the time that arbitration modules 3 detects that cache module 2 has the data of transmission, data head to the interior transmission data of cache module 2 is resolved, when send data all deposit in cache module 2 interior after, arbitration modules 3 selects corresponding network interface transceiving data module to carry out data transmission according to sending the data head of data; In the time that network interface transceiving data module detects network feedback data, packet number and the byte number of sending/receiving module statistics network feedback data, and network feedback data is all deposited in the memory module being connected with described sending/receiving module together with packet number and the byte number of statistics; Arbitration modules 3 reads the data of storage in memory module and adds required data head and receives data to form network-feedback, and described network-feedback reception data are deposited in cache module 2, reads for PCIEX8 EBI 1.
Particularly, cache module 2 circulate between PCIEX8 EBI 1 and arbitration modules 3 for data need memory block, PCIEX8 EBI 1 is existing conventional a kind of EBI form. In the embodiment of the present invention, storage device 4 at least comprises two memory modules, has illustrated that storage device 4 comprises four memory modules, i.e. the first memory module, the second memory module, the 3rd memory module and the 4th memory module in Fig. 1. In the time that storage device 4 comprises four memory modules, in transceiver 5, comprise four sending/receiving modules, meanwhile, in network interface transceiving data device 6, comprise four network interface transceiving data modules (MAC); Connect in order to realize one to one, the first memory module is connected with the first sending/receiving module, the first sending/receiving module is connected with the first network interface transceiving data module, the second memory module is connected with the second sending/receiving module, the second sending/receiving module is connected with the second network interface transceiving data module, the 3rd memory module is connected with the 3rd sending/receiving module, the 3rd sending/receiving module is connected with the 3rd network interface transceiving data module, the 4th memory module is connected with the 4th sending/receiving module, and the 4th sending/receiving module is connected with the 4th network interface transceiving data module. In the specific implementation, the interior memory module of storage device 4 can, more than four, send the quantity comprising in reception 5 and network interface transceiving data device 6 corresponding consistent with memory module, can meet the needs that connect one to one.
Further, arbitration modules 3 is numbered the network interface transceiving data module in network interface transceiving data device 6, in the time carrying out data transmission, can select to send by the network interface transceiving data module of reference numeral, in the time carrying out data receiver, can add data head according to the numbering of network interface transceiving data module. Multiple network interface transceiving data modules that the present invention connects by arbitration modules 3 are carried out the operation of data transmit-receive simultaneously, realize the function that parallel data sends.
Described arbitration modules 3 is numbered the network interface transceiving data module in network interface transceiving data device 6, and the data head that arbitration modules 3 is added comprises the numbering that is connected network interface transceiving data module with memory module, packet number and the byte number of network feedback data.
Described PCIEX8 EBI 1 is connected with host computer, and the network-feedback that host computer 1 reads in cache module 2 by PCIEX8 EBI 1 receives data, and can transmit required transmission data to PCIEX8 EBI 1.
As shown in Figure 2, Figure 3 and Figure 4: a kind of based on parallel pipeline network data processing method, described network data processing method comprises the steps:
A, receive and send data by PCIEX8 EBI 1, and described transmission data are deposited in cache module 2;
The transmission data of PCIEX8 EBI 1 are that host computer sends, and the transmission data of host computer transmission have been specified and need to have been sent by the network interface transceiving data module of corresponding numbering.
B, arbitration modules 3 detect in real time to cache module 2, and in the time having data in cache module 2, arbitration modules 3 is resolved the data head of the interior transmission data of cache module 2; Arbitration modules 3 judge according to sending the data head of data whether send data enters cache module 2 completely, when send data enter completely cache module 2 interior after, arbitration modules 3 deposits transmission data in corresponding memory module according to the data head of transmission data;
In described step b, arbitration modules 3 is resolved the numbering of the network interface transceiving data module that obtains sending packet number, the byte number of data and outwards send to sending the data head of data. Arbitration modules 3 can judge according to the packet number of resolving and byte number whether send data enters in cache module 2 completely. Obtain sending in parsing after the numbering of the network interface transceiving data module of setting, arbitration modules 3 deposits transmission data with network interface transceiving data module in and numbers in corresponding memory module, while transmission as passed through the first network interface transceiving data module, need to deposit in the first memory module sending data.
C, the sending/receiving module being connected with memory module detect and send after data, and sending/receiving module will send data and will send data by network interface transceiving data module and outwards send;
Further, when arbitration modules 3 is resolved sending the data head of data, can also obtain the number of times that described transmission data need to send, in the time sending by network interface transceiving data module, also need to detect the number of times sending and whether reach the transmission times of setting in the data head that sends data, in the time not arriving transmission times, need to repeat to send, until meet the transmission times of setting, thereby realize the object that streamline sends.
D, by network interface transceiving data module receive network feedback data, sending/receiving module is added up packet number, the byte number of network feedback data, and by network feedback data together with statistics packet number, byte number all deposit in memory module;
E, arbitration modules 3 read network feedback data, packet number, the byte number in memory module, and according to the numbering of the network interface transceiving data module being connected with memory module, network feedback data are added to data head, receive data to form network-feedback;
When having while all having network feedback data in multiple memory modules, arbitration modules 3 reads successively according to the order of network interface transceiving data module numbering. In the embodiment of the present invention, described order reads the order referring to according to network interface transceiving data module is numbered, as read successively according to the order of the first network interface transceiving data module, the second network interface transceiving data module, the 3rd network interface transceiving data module and the 4th network interface transceiving data module.
F, arbitration modules 3 receive data by network-feedback and deposit in cache module 2, read described network-feedback receive data for PCIEX8 EBI 1.
The present invention receives transmission data by PCIEX8 EBI 1, arbitration modules 3 receives sending data, judgement and parsing, and send by corresponding network interface transceiving data module, the network feedback data that arbitration modules 3 receives network interface transceiving data module is added data head and is formed network-feedback reception data, network-feedback is received to data to be deposited in cache module 2, after reading for PCIEX8 EBI 1, be transferred to host computer, its compact conformation, can realize the network data processing of parallel pipeline, improve the speed of data processing, wide accommodation, safe and reliable.
Claims (6)
1. one kind based on parallel pipeline network data processing device, it is characterized in that: comprise PCIEX8 EBI (1), described PCIEX8 EBI (1) is connected with arbitration modules (3) by cache module (2), described arbitration modules (3) is connected with transceiver (5) by storage device (4), described transceiver (5) is connected with network interface transceiving data device (6), described storage device (4) comprises at least two memory modules, in transceiver (5), comprise and the corresponding consistent sending/receiving module of storage module number in storage device (4), in network interface transceiving data device (6), comprise and the corresponding consistent network interface transceiving data module of memory module quantity, sending/receiving module in memory module and transceiver (5) in storage device (4) connects one to one, and memory module connects one to one by sending/receiving module and network interface transceiving data module,
PCIEX8 EBI (1) deposits the transmission data that receive in cache module (2) in, in the time that arbitration modules (3) detects that cache module (2) has the data of transmission, the data head that sends data in cache module (2) is resolved, after transmission data all deposit in cache module (2), arbitration modules (3) selects corresponding network interface transceiving data module to carry out data transmission according to the data head that sends data; In the time that network interface transceiving data module detects network feedback data, packet number and the byte number of sending/receiving module statistics network feedback data, and network feedback data is all deposited in the memory module being connected with described sending/receiving module together with packet number and the byte number of statistics; Arbitration modules (3) reads the data of storage in memory module and adds required data head and receives data to form network-feedback, and described network-feedback reception data are deposited in cache module (2), reads for PCIEX8 EBI (1).
2. according to claim 1 based on parallel pipeline network data processing device, it is characterized in that: described arbitration modules (3) is numbered the network interface transceiving data module in network interface transceiving data device (6), the data head that arbitration modules (3) is added comprises the numbering that is connected network interface transceiving data module with memory module, packet number and the byte number of network feedback data.
3. according to claim 1 based on parallel pipeline network data processing device, it is characterized in that: described PCIEX8 EBI (1) is connected with host computer, the network-feedback that host computer reads in cache module (2) by PCIEX8 EBI (1) receives data, and can transmit required transmission data to PCIEX8 EBI (1).
4. based on a parallel pipeline network data processing method, it is characterized in that, described network data processing method comprises the steps:
(a), receive and send data by PCIEX8 EBI (1), and described transmission data are deposited in cache module (2);
(b), arbitration modules (3) detects in real time to cache module (2), in the time having data in cache module (2), arbitration modules (3) is resolved the data head that sends data in cache module (2); Arbitration modules (3) judges according to the data head that sends data whether send data enters cache module (2) completely, after transmission data enter in cache module (2) completely, arbitration modules (3) deposits transmission data in corresponding memory module according to the data head that sends data;
(c), the sending/receiving module that is connected with memory module detects and sends after data, sending/receiving module will send data and will send data by network interface transceiving data module and outwards send;
(d), receive network feedback data by network interface transceiving data module, sending/receiving module is added up packet number, the byte number of network feedback data, and network feedback data is all deposited in memory module together with packet number, the byte number of statistics;
(e), arbitration modules (3) reads network feedback data, packet number, the byte number in memory module, and according to the numbering of the network interface transceiving data module being connected with memory module, network feedback data is added to data head, receive data to form network-feedback;
(f), arbitration modules (3) receives data by network-feedback and deposits in cache module (2), read described network-feedback for PCIEX8 EBI (1) and receive data.
5. according to claim 4ly it is characterized in that based on parallel pipeline network data processing method, when having while all having network feedback data in multiple memory modules, arbitration modules (3) reads successively according to the order of network interface transceiving data module numbering.
6. according to claim 4 based on parallel pipeline network data processing method, it is characterized in that, in described step (b), arbitration modules (3) is resolved the numbering of the network interface transceiving data module that obtains sending packet number, the byte number of data and outwards send to sending the data head of data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310662519.1A CN103678235B (en) | 2013-12-09 | 2013-12-09 | Based on parallel pipeline network data processing device and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310662519.1A CN103678235B (en) | 2013-12-09 | 2013-12-09 | Based on parallel pipeline network data processing device and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103678235A CN103678235A (en) | 2014-03-26 |
| CN103678235B true CN103678235B (en) | 2016-06-08 |
Family
ID=50315851
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310662519.1A Active CN103678235B (en) | 2013-12-09 | 2013-12-09 | Based on parallel pipeline network data processing device and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103678235B (en) |
Families Citing this family (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9753876B1 (en) | 2016-09-22 | 2017-09-05 | International Business Machines Corporation | Processing of inbound back-to-back completions in a communication system |
| CN109687943B (en) * | 2018-10-11 | 2021-10-12 | 中国人民解放军海军陆战队训练基地 | Dual-computer backup redundancy control system |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102394732A (en) * | 2011-09-06 | 2012-03-28 | 中国人民解放军国防科学技术大学 | Multi-micropacket parallel processing structure |
| CN102650976A (en) * | 2012-04-01 | 2012-08-29 | 中国科学院计算技术研究所 | Control device and method supporting single IO (Input/Output) virtual user level interface |
| CN102855090A (en) * | 2012-07-23 | 2013-01-02 | 深圳市江波龙电子有限公司 | Storage equipment and running method thereof |
| CN103109260A (en) * | 2010-05-04 | 2013-05-15 | 谷歌公司 | Parallel processing of data |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7822889B2 (en) * | 2007-08-27 | 2010-10-26 | International Business Machines Corporation | Direct/indirect transmission of information using a multi-tiered full-graph interconnect architecture |
-
2013
- 2013-12-09 CN CN201310662519.1A patent/CN103678235B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103109260A (en) * | 2010-05-04 | 2013-05-15 | 谷歌公司 | Parallel processing of data |
| CN102394732A (en) * | 2011-09-06 | 2012-03-28 | 中国人民解放军国防科学技术大学 | Multi-micropacket parallel processing structure |
| CN102650976A (en) * | 2012-04-01 | 2012-08-29 | 中国科学院计算技术研究所 | Control device and method supporting single IO (Input/Output) virtual user level interface |
| CN102855090A (en) * | 2012-07-23 | 2013-01-02 | 深圳市江波龙电子有限公司 | Storage equipment and running method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103678235A (en) | 2014-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN202261360U (en) | Device for testing robustness of router data path | |
| CN103714029B (en) | Novel two-line synchronous communication protocol and application | |
| CN103490959B (en) | A kind of dual-redundant CAN bus fault detection method | |
| CN103178872B (en) | Method and the device of USB system transfers distance is extended by Ethernet | |
| CN104135741A (en) | Method of Bluetooth low energy communication, Bluetooth devices and system | |
| CN104199795B (en) | Data transferring and receiving method of bus framework | |
| CN104572574A (en) | GigE (gigabit Ethernet) vision protocol-based Ethernet controller IP (Internet protocol) core and method | |
| CN104281548A (en) | Method, device and system for data transmission based on AXI bus | |
| CN1812317A (en) | Synchronous medium access controller | |
| CN104750588A (en) | Serial port communication based pressure testing method | |
| CN103744811A (en) | Serial data transmission system and method | |
| CN102420763A (en) | Direct memory access (DMA) sending method | |
| CN103678235B (en) | Based on parallel pipeline network data processing device and method | |
| CN104731737A (en) | Hub, operating system and control method | |
| CN204256732U (en) | The high-speed data transmission apparatus of Based PC I-Express interface | |
| CN103107862B (en) | Logical device and MDIO data transmission method for uplink thereof | |
| CN105356988A (en) | PCIe based full duplex DMA transmission method | |
| CN106027397A (en) | Network communication method for distributed measuring equipment of star expansion | |
| CN103106177B (en) | Interconnect architecture and method thereof on the sheet of multi-core network processor | |
| CN203982361U (en) | A kind of multiple serial communication proving installation | |
| CN102497296A (en) | Method for testing normality of network card receiving and sending packets | |
| CN102447597B (en) | Method and device for realizing IP (Internet Protocol) effectiveness detection | |
| CN204597989U (en) | Based on the AFDX terminal test equipment of CPCI interface | |
| CN105897494A (en) | Testing method of network transmission rate | |
| CN110096002A (en) | A kind of automatization test system and test method based on CANFD bus |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant |