CN103440200B - A kind of height based on dual operating systems real-time big data quantity test back method - Google Patents
A kind of height based on dual operating systems real-time big data quantity test back method Download PDFInfo
- Publication number
- CN103440200B CN103440200B CN201310399154.8A CN201310399154A CN103440200B CN 103440200 B CN103440200 B CN 103440200B CN 201310399154 A CN201310399154 A CN 201310399154A CN 103440200 B CN103440200 B CN 103440200B
- Authority
- CN
- China
- Prior art keywords
- data
- node
- buffer area
- variable
- file
- 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
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
A kind of height based on dual operating systems real-time big data quantity test back method, solve a difficult problem for the high test of big data quantity in real time playback, first data produce node request for data buffer area, data consumer node request for data buffer area, all buffer areas of initialization, look for data file and obtain data in buffer area, initialization network connects, data consumption node produces node requests data to data, data produce node to data consumption node return data, and data consumption node requires to peek from buffer memory according to task scheduling.The present invention effectively can avoid the excessive problem causing file transmission time long of data file, and along with the increase of data file, test macro can not obviously increase for the expense of testing playback.In addition, by shortening the length of each communication data packet, multiple variable also effectively can be avoided simultaneously to need the network congestion caused when transmitting data, improve the reliability of system.
Description
Technical field
The present invention relates to a kind of height based on dual operating systems real-time big data quantity test back method, belong to embedded software test field.
Background technology
In common test macro, test case generally realizes in the mode of script, or directly realizes with the form of code.But in the test process of reality, tester needs data to change in the mode of serializing sometimes, such as emulates flight path or sets up realistic model etc. according to the temperature feature in area a certain in certain a period of time.When structure test environment, this demand is especially common, such as, in certain system under test (SUT), assembly 1 sends data to assembly 2, the data that assembly 2 sends according to assembly 1 make corresponding feedback, when the test environment building assembly 2, traditional method requires that the function and structure of tester to assembly 1 has suitable understanding, must the course of work of simulated assembly 1, but simulated assembly 1 course of work will consume a large amount of time and manpower and materials, and the reasons such as deficiency may be coordinated to cause the work of simulated assembly 1 to prove an abortion due to privacy requirements and exploitation side.If the data that assembly 1 sends under the various duty of various working stage can be preserved and send to assembly 2 by test macro simulated assembly 1, then just can build the semi-hardware type simulation test environment of assembly 2 very soon.
For the data of this irregular change, or cannot with the data of algorithm simulation change because realize the reasons such as difficulty is large, operational efficiency is low, a kind of method must be provided, test data in previous test process can be preserved, and these data can be exported to system under test (SUT) in real time.
Current embedded software not only function is increasingly sophisticated, performance is also more and more higher, the cycle that much software carries out message exchange can reach Millisecond and even reach Microsecond grade, therefore requires that the test macro of the embedded software minimum cycle of operation can reach Millisecond and even Microsecond grade; Meanwhile, the data volume also making to produce in test process that communicates frequently increases, and communication cycle is little is greatly that high big data quantity in real time tests the difficult point of playback with data volume.
Adopt Windows system easy and simple to handle as the test macro of running environment, friendly interface, easily test data can be saved on disk and in test process, to open data file and export, but real-time is very poor, can not meet the demand of embedded software test, vxworks operating system is a kind of real time operating system, real-time can meet the test of Millisecond and even Microsecond grade completely, but vxworks operating system interface is friendly not, the more important thing is as a kind of real time embedded system, VxWorks system is not suitable for operating fixed disk file, especially the data file of big data quantity, for reaching the object of test playback, the data file acquisition content wherein opened in hard disk is needed to be sent by bus interface again, if adopt disposable file reading content to the method in internal memory, need to consume very large one piece of memory headroom, even cause EMS memory error, system crash, relative, if read data files several times, only read in a part of content of regular length at every turn, need again to operate hard disk continually, be easy to the operation affecting real-time task.
Summary of the invention
Technology of the present invention is dealt with problems: overcome the deficiencies in the prior art, a kind of height based on dual operating systems real-time big data quantity test back method is provided, when implementing test, data file can be sent from bus interface, also supporting that the data file of user being write voluntarily is as the source of data; Support that the cycle of operation reaches the test of Millisecond, the hard disk of VxWorks system or other real time operating systems can not be operated frequently, can not increase significantly along with data volume increases the memory source testing playback consumption in addition.
The technology of the present invention solution: a kind of height based on dual operating systems real-time big data quantity test back method, the problem that first will solve is the source of data, by the support to common data form, user can manually write, and the various method such as algorithm generation generates data file; Test macro generally all has the ability of collecting test data, and by specifying a kind of unified form, the test data of test macro collection can generate data file easily and use as Data Source in test activity afterwards.The test macro applied along with this method of concrete file layout is different may be different, but to store data in the matrix form in data file be basic key element.
For reaching high real-time testing replay request, the present invention must obtain data in a short period of time and send from bus interface from data file, the strict task of this part requirement of real-time must run on real time operating system, such as VxWorks, real-time linux etc., because the work period of test activity is very little, therefore data volume in data file is often very large, disposable for the data file internal memory that reads in can be consumed too much memory source, so data file should be divided into the data block of fixed size, each reading also processes the data of certain length, but so just require to operate hard disk frequently, the execution of real-time task may be affected, this part task is to real-time not requirement, therefore can place it in windows operating system and carry out, the test data that a lot of test macro collections is returned on the one hand all carries out treatment and analysis in windows system, directly these test datas can be exported as data file, user also directly easily can write data file on windows on the other hand.Non-real time operating system and the node of service data file will be run according to role to be called " data production node ", will real time operating system be run and the node that real operator trunk interface sends test data is called " data consumption node ".
The present invention obtains data and sends so that the implementation procedure realizing test playback is as follows from bus interface from data file:
(1) data produce node needs the buffer area of application space according to the length computation of variable corresponding to data file, and concrete computing method are (length once reading the number k* variable of data in buffer area).Need simultaneously from data file, to obtain data if there is multiple variable in test macro, then need to be each variable application buffer area according to the length of each variable respectively.Each number k reading data in buffer area produces node memory condition according to data and needs the length of every bar data in the buffer area quantity of application and buffer area to determine, internal memory is larger, suitably can increase each number read, and needs to apply for that the length of every bar data in the many or buffer area of the quantity of buffer area is long and then reduces each number read to reduce the consumption to internal memory.
(2) variable that data consumption node traverses is all, statistics needs from data file, obtain data and carries out testing the information such as the length of the variable number of playback and the ID of variable and variable.For each variable application one group of buffer area is used for the data in store data file, often organize buffer area and comprise the same buffer area of n block size, the size of buffer area is the number of plies that the length of variable is multiplied by the data of preserving in buffer area, here the data in the number of plies of indication and buffer area can for the number of times of variable consumption, and can obtain is once one deck.The buffer area of each variable on data consumption node is made up of n block cache district, and to avoid read/write conflict, the size of n depends on the speed of operating system read/write memory, and faster then n is less for speed.
(3) all buffer area initialization record start position, the information such as the position of buffer size and operation at present, each needs oneself current operation of variable record of production data to data consumption node is which block cache.The corresponding read-write state label of each buffer area of data consumption node, for the state read and write data in record buffer memory district, so that the data consumption in buffer area makes corresponding process after completing.
(4) open data file and obtain data in buffer area, record data are produced node and from each data file, are obtained the number of times of data to calculate the position that next time obtains data from file, concrete computing method are (number of times once reading the number k* acquisition data of data in buffer area), when opening certain data file at every turn, from this position, obtain the buffer area of k bar data (k is the number once reading data in buffer area) stored in variable corresponding to this data file, if the data deficiencies in file, in buffer area, insert end mark.
(5) data are produced node and are entered listening state as data delivery service device end, and data consumption node binds the port connection server end of arranging in advance as client.
(6) data consumption node is produced node to data and is sent message, require that data are produced node and started production data, the content of message comprises the ID that request generates the variable of data, and the number of plies of request msg and the data of asking will stored in which block cache districts, data are produced after node receives message and in the data buffer area belonging to this variable, are obtained data according to the ID of variable in message, if the data in buffer area are less than the number of plies of data consumption node request, then repeat step 2, data are obtained from file, if have end mark to exist in the data sent, illustrate that the data in buffer memory have been not enough to meet the request of data consumption node, then only send valid data to data consumption node, before sending data, shake hands by sending out message, the content of message comprise this ID of variable belonging to the data sent out, this the length of packet sent out, and the data in packet should stored in which block cache.
(7) data consumption node is produced node institute according to data and is sent out the processing mode that message judges to receive after data, and mainly comprising data should stored in which block cache, and whether data length is satisfied asks.Data are produced the data that the handshake information that sends over of node comprises in the ID of variable belonging to data and packet and stored in which block cache of variable, should can be obtained primary data should stored in which block cache according to these two information.If data to produce the number of data layers that returns of node consistent with the number of plies of request, then by data stored in wait for scheduler task acquisition data in buffer memory and assignment to variable; If the number of data layers that data production node returns is greater than the number of plies of request, then illustrate that mistake appears in communication process, provides warning; If the number of data layers that data production node returns is less than the number of plies of request, then illustrate that the data in data file are finished using, data consumption node no longer produces node requests data to data, stops data readback operating after the data consumption in two block cache districts.
(8) data consumption node obtains data according to the requirement of dispatching algorithm from buffer memory, obtains one deck at every turn,
After the data in a block cache district are all acquired, judge what operation data consumption node should make according to the value of the read-write state label of buffer area: if the data length of this buffer area acquisition last time and the identical of request, illustrate that data are produced node and can also be produced enough data; If the data of this buffer area acquisition last time are less than the data of request, then illustrate that data are produced node and can not be produced enough data, then the data in the buffer area belonging to this variable have all been consumed rear stopping data readback operation; If the data length of this buffer area acquisition last time is greater than the length of request, illustrative system runs makes mistakes, and provides and warns and stop testing playback operation.
The present invention's advantage is compared with prior art: the invention provides the difficult problem that a kind of method solves the high test of big data quantity in real time playback, produce node and data consumption node two kinds of roles by dividing data and give different responsibilities, operation does not require a great deal of time with the data consumption node on real time operating system and opens disk file, the index safeguarding file reading is not needed yet, as long as the requirement according to oneself is peeked from buffer area, thus fully played the advantage of real time operating system, make to adopt test macro of the present invention can support the test playback of high real-time embedded software, take layering to transmit the mode of data, each reading finite layer data, effectively can avoid the excessive problem causing file transmission time long of data file, along with the increase of data file, test macro can not obviously increase for the expense of testing playback simultaneously.In addition, by shortening the length of each communication data packet, multiple variable also effectively can be avoided simultaneously to need the network congestion caused when transmitting data, improve the reliability of system.
Accompanying drawing explanation
Fig. 1 is the schematic flow sheet of the height real-time big data quantity test back method based on dual operating systems;
Fig. 2 is the data batchmove schematic diagram in the inventive method;
Fig. 3 is the implementation procedure that in the inventive method, data are sent to data consumption node from data file;
Fig. 4 is that in the inventive method, data produce the implementation procedure operated after node receives request of data;
Fig. 5 is the implementation procedure that in the inventive method, data consumption node operates after receiving data;
Embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearly, be described in detail below in conjunction with the accompanying drawings and the specific embodiments.
The object of the invention is to overcome the deficiencies in the prior art, a kind of height based on dual operating systems real-time big data quantity test back method is provided, pass through the present invention, the test data collected in test process can be exported as data file by test macro, when implementing test, data file being sent from bus interface, also can support that the data file of user being write voluntarily is as the source of data.The present invention must support that the cycle of operation reaches the test of Millisecond, can not operate the hard disk of VxWorks system or other real time operating systems frequently, should not increase significantly in addition along with data volume increases the memory source testing playback consumption.
Introduce the present invention's application in a particular embodiment for clearer, first to embodiment for the structure of system do simple introduction:
This test macro is real-time testing system, real-time task run with VxWorks system on, to the not high task run of requirement of real-time on Windows system, such as man-machine interaction, test data analysis etc., the test data produced in test process can stored in MySql database after test terminates.
First for realizing test playback, the test data that test macro produces must to be re-entered in a certain format in test macro and the variable in test macro is changed according to the change of data, and this just requires that test macro must possess ability test result being exported as data file.
Test system and test result in the present embodiment is kept in MySql database, by selecting the configurable variable needing to derive of the row of the data file derived, generate different SQL statement and obtain result from database, then with a matrix type result is outputted in text, wherein, the result of same variable is stored in same row, uses tab segmentation in the middle of row.For other test macros, as long as unified the form of the data file derived, method of the present invention can be adopted to carry out test playback, and the preservation form of the framework of test macro itself, actualizing technology and test result is all irrelevant.User also manually can write or generate data file with algorithm, as long as according to form affiliated above, can be applied in this test back method.
As shown in Figure 1, the present invention's concrete steps are in an embodiment as follows:
Step one: first data produce node should apply in internal memory that buffer area is used for depositing the data obtained from data file, multiple needs may be there are in system simultaneously and from data file, obtain the variable that data carry out testing playback, so data file corresponding to each variable must be recorded.Test macro data in embodiment are produced on node and are used class to manage the attribute information of variable, and the class for variable increases a character string member, for the data file that record variable is corresponding; Test macro of the present invention is applied for other, if original system adopts the attribute of class or structure record variable, then increase data file corresponding to member record variable wherein, if original system does not have similar structure to manage variable's attribute, array can be used variable ID and data file one_to_one corresponding and do with record.
In test macro, the length of each variable may be different, and need the space of the buffer area of application according to the length computation of variable corresponding to data file, concrete computing method are (length once reading the number k* variable of data in buffer area).Need simultaneously from data file, to obtain data if there is multiple variable in test macro, then need to be each variable application buffer area according to the length of each variable respectively.Each number k reading data in buffer area produces node memory condition according to data and needs the length of every bar data in the buffer area quantity of application and buffer area to determine, internal memory is larger, suitably can increase each number read, and needs to apply for that the length of every bar data in the many or buffer area of the quantity of buffer area is long and then reduces each number read to reduce the consumption to internal memory.In the present embodiment, every bar data length is identical with corresponding variable-length, is 1024 bytes to the maximum, and the variable number of test macro support is 300, therefore in the embodiment of the present invention, k is decided to be 1000 and can ensures that system has enough internal memories to complete data readback task.
Step 2: data consumption node produces from data the various attribute informations that node obtains variable, travels through all variablees, statistics needs from data file, obtain data and carries out testing the information such as the length of the variable number of playback and the ID of variable and variable.For each variable application one group of buffer area is used for the data that store data produces node generation, often organize buffer area and comprise the same buffer area of n block size, the size of buffer area is the number of plies that the length of variable is multiplied by the data of preserving in buffer area, here the data in the number of plies of indication and buffer area can for the number of times of variable consumption, and can obtain is once one deck.Produce node for avoiding data to send the data window phase to data consumption node and longly cause network congestion, the number of plies of each acquisition data is too much unsuitable, in embodiments of the invention, each number of plies obtaining data is 20 layers, and the maximum variable-length of system support is 1024 bytes, such acquisition one secondary data maximum length is 20kB, concerning the VxWorks system supporting multitask real time execution, the data consumes receiving and process 20kB is very little, even if carrying out testing with the cycle of ten Microsecond grades is also the data transmitting 20kB in needs 200 microsecond, the operation of real-time task can not be affected, the load caused network is also very little.In the embodiment of the present invention, real-time simulation node runs on real time operating system VxWorks, and the speed of copies data is very fast, therefore each variable uses two block cache district alternately preserves data.
In the present invention, data produce the visible Fig. 2 of data transfer process between node and data consumption node.
In the present invention, data produce visible Fig. 3 of implementation procedure of node production data and data consumption node consumption data.
Step 3: all buffer area initialization record start position, the information such as the position of buffer size and operation at present, each needs oneself current operation of variable record of production data to data consumption node is which block cache.In an embodiment of the present invention, use memset function that all buffer areas are initialized as 0, what the starting position, at present operating position, buffer size and each variable that use a structure to manage all buffer areas operated at present is which block cache of oneself etc. information.The corresponding read-write state label dataRWLabel of each buffer area of data consumption node, for the state that record buffer memory district reads and writes data, so that the data consumption in buffer area makes corresponding process after completing.
Step 4: open data file and obtain data in buffer area, record data are produced node and from each data file, are obtained the number of times of data to calculate the position that next time obtains data from file, concrete computing method are (number of times that 1000* obtains data), when opening certain data file at every turn, from this position, obtain the buffer area of 1000 data stored in variable corresponding to this data file, if the data deficiencies in file, in buffer area, insert end mark, in embodiment, use " $ " as end mark.
Step 5: data are produced node and entered listening state as data delivery service device end, the data consumption node variable that to be each needs obtain data from network opens a client and the port of binding agreement is in advance connected to server end.
Step 6: data readback produces node requests data for starting point with data consumption node to data, in an embodiment of the present invention, test macro must operate through test initialization before the test begins, produce node requests data when test initialization carries out to data two block cache districts of each variable to be filled up data and can reduce consumption to system resource when test is carried out, ensure to test in the buffer area of data consumption node at the beginning namely have enough data to carry out data readback operation.Data consumption node is produced node to data and is sent message, and require that data are produced node and started production data, the content of message comprises the ID that request generates the variable of data, and the number of plies of request msg and the data of asking will stored in which block cache districts; Data are produced after node receives message and in the data buffer area belonging to this variable, are obtained data according to the ID of variable in message, if remaining data are less than the number of plies of data consumption node request in buffer area, then repeat step 2, from file, obtain data.The data first checked before data in buffer area are sent to data consumption node, if have end mark to exist in the data sent, illustrate that the data in buffer area have been not enough to meet the request of data consumption node, then only send valid data to data consumption node.Sending before data, shaking hands by sending out message, the content of message comprise this ID of variable belonging to the data sent out, this length of packet sent out, and the data in packet should stored in which block cache.Data produce visible Fig. 4 of implementation procedure that node receives the operation after the data request information of data consumption node.
Step 7: data consumption node is produced node institute according to data and sent out message judgement and receive the processing mode after data, visible Fig. 5 of implementation procedure of this step.Mainly comprising data should stored in which block cache, and whether data length meets request etc.Data are produced the data that the handshake information that sends over of node comprises in the ID of variable belonging to data and packet and stored in which block cache of variable, should can be obtained primary data should stored in which block cache according to these two information.
The data package size needing to receive can be determined by the data packet length in handshake information, for ensureing real-time, the work receiving packet from network should the operation of influential system task scheduling, therefore should run on independently in thread, in the present embodiment, the system that data consumption node runs is VxWorks system, therefore the work receiving packet from network should run on independently task, when initialization buffer area, each needs call taskSpawn(from the variable of Network Capture data) function produces a task, the priority of this task should lower than the priority of system call task, the entrance function circulation of task obtains semaphore, obtain and be then in idling conditions less than semaphore, if obtain handshake information from network, entering duty, call fioRead() function obtains data from network.Function described in literary composition is VxWorks system interface function, if the present invention is applied to other system, then the function called may be different with operation, but concrete algorithm can be continued to use.
The data received from network need to judge that its length decides subsequent operation with whether the length of request is consistent, by composing with different values to buffer area read-write state label dataRWLabel, can judge that data produce the duty of node according to the value of this label in subsequent operation.If data to produce the number of data layers that returns of node consistent with the number of plies of request, then by data stored in wait for scheduler task acquisition data in buffer memory and assignment to variable, be 1 by dataRWLabel assignment; If the number of data layers that data production node returns is greater than the number of plies of request, then illustrate that mistake appears in communication process, providing warning, is 44 by dataRWLabel assignment; If the number of data layers that data production node returns is less than the number of plies of request, then illustrate that the data in data file are finished using, data consumption node no longer produces node requests data to data, data readback is stopped to operate after data consumption in two block cache districts, being 88 by dataRWLabel assignment corresponding for current cache district, is 66 by dataRWLabel assignment corresponding for another block cache.
Step 8: data consumption node obtains data according to the requirement of dispatching algorithm from buffer memory, each acquisition one deck, after the data in a block cache district are all acquired, judge what operation data consumption node should make according to the value of the read-write state label dataRWLabel of buffer area:
1) if the data length of this buffer area acquisition last time and the identical of request, that is dataRWLabel==1, illustrate data produce node can also produce enough data, to data produce node initiate a message request msg to current buffer area and start from another block cache reading data;
2) if the data of this buffer area acquisition last time are less than the data of request, that is dataRWLabel==88, then illustrate that data are produced node and can not be produced enough data, data then in this buffer area no longer produce node requests data to data after all consuming, and the data in this buffer area have all been consumed rear stopping data readback operation;
3) if the data length of this buffer area acquisition last time is greater than the length of request, illustrative system runs makes mistakes, and provides and warns and stop testing playback operation.
4) if dataRWLabel==66, illustrate that data are produced node and can not have been continued to produce enough data, and another buffer area obtains ultimogenitary data, data in this buffer area no longer produce node requests data to data after all consuming, by complete for the data consumption in all buffer areas rear stopping data readback operation.
Data file has been consumed and test process stops all causing data readback to stop, and needs in this process to carry out following work:
1) all data buffer area are all emptied, all pointed NULL, if application buffer area uses dynamic application, then must the space of release dynamics application.
2) stop and deleting all tasks for obtaining data from network, calling KERNEL_taskDelete(in the present embodiment) function complete this work.
3) disconnecting each client to the connection of server end closes client and data and produces server end on node.
Non-elaborated part of the present invention belongs to techniques well known.
The above; be only part embodiment of the present invention, but protection scope of the present invention is not limited thereto, any those skilled in the art are in the technical scope that the present invention discloses; the change that can expect easily or replacement, all should be encompassed within protection scope of the present invention.
Claims (1)
1. the real-time big data quantity test of the height based on a dual operating systems back method, it is characterized in that: non-real time operating system and the node of service data file will be run and be called that data produce node, to run the real time operating system node that also real operator trunk interface sends test data and be called data consumption node, performing step is as follows:
(1) data produce node needs the buffer area of application space according to the length computation of variable corresponding to data file, and concrete computing method are once read the number k of data in buffer area to be multiplied by the length of variable; Need simultaneously from data file, to obtain data if there is multiple variable in test macro, then need to be each variable application buffer area according to the length of each variable respectively; Each number k reading data in buffer area produces node memory condition according to data and needs the length of every bar data in the buffer area quantity of application and buffer area to determine, internal memory greatly then increases each number read, and needs to apply for that the length of every bar data in the many or buffer area of the quantity of buffer area is long and then reduces each number read to reduce the consumption to internal memory;
(2) variable that data consumption node traverses is all, statistics needs from data file, obtain the length information that data carry out testing the variable number of playback and the ID of variable and variable, for each variable application one group of buffer area is used for the data in store data file, often organize buffer area and comprise the same buffer area of n block size, the size of buffer area is the number of plies that the length of variable is multiplied by the data of preserving in buffer area, data in the described number of plies and buffer area can for the number of times of variable consumption, and can obtain is once one deck; The buffer area of each variable on data consumption node is made up of n block cache district, and to avoid read/write conflict, the size of n depends on the speed of operating system read/write memory, and faster then n is less for speed;
(3) all buffer area initialization record start position, the positional information of buffer size and operation at present, which block cache what in data consumption node, each needed oneself current operation of variable record of production data is; The corresponding read-write state label of each buffer area of data consumption node, for the state read and write data in record buffer memory district, so that the data consumption in buffer area makes corresponding process after completing;
(4) open data file and obtain data in buffer area, record data are produced node and from each data file, are obtained the number of times of data to calculate the position that next time obtains data from file, concrete computing method are once read the number k of data in buffer area to be multiplied by the number of times obtaining data, when opening certain data file at every turn, from this position, obtain k bar data, by the buffer area of k number stored in variable corresponding to this data file, if the data deficiencies in file, in buffer area, insert end mark;
(5) data are produced node and are entered listening state as data delivery service device end, and data consumption node binds the port connection server end of arranging in advance as client;
(6) data consumption node is produced node to data and is sent message, require that data are produced node and started production data, the content of message comprises the ID that request generates the variable of data, and the number of plies of request msg and the data of asking will stored in which block cache districts, data are produced after node receives message and in the data buffer area belonging to this variable, are obtained data according to the ID of variable in message, if the data in buffer area are less than the number of plies of data consumption node request, then repeat step (2), data are obtained from file, if have end mark to exist in the data sent, illustrate that the data in buffer memory have been not enough to meet the request of data consumption node, then only send valid data to data consumption node, before sending data, shake hands by sending out message, the content of message comprise this ID of variable belonging to the data sent out, this the data in the length of packet sent out and packet should stored in any block cache,
(7) data consumption node is produced node institute according to data and is sent out the processing mode that message judges to receive after data, and mainly comprising data should stored in which block cache, and whether data length is satisfied asks; Data are produced the data that the handshake information that sends over of node comprises in the ID of variable belonging to data and packet and stored in which block cache of variable, should can be obtained primary data should stored in which block cache according to these two information; If data to produce the number of data layers that returns of node consistent with the number of plies of request, then by data stored in wait for scheduler task acquisition data in buffer memory and assignment to variable; If the number of data layers that data production node returns is greater than the number of plies of request, then illustrate that mistake appears in communication process, provides warning; If the number of data layers that data production node returns is less than the number of plies of request, then illustrate that the data in data file are finished using, data consumption node no longer produces node requests data to data, stops data readback operating after the data consumption in two block cache districts;
(8) data consumption node obtains data according to the requirement of dispatching algorithm from buffer memory, each acquisition one deck, after the data in a block cache district are all acquired, judge what operation data consumption node should make according to the value of the read-write state label of buffer area: if the data length of this buffer area acquisition last time and the identical of request, illustrate that data are produced node and can also be produced enough data; If the data of this buffer area acquisition last time are less than the data of request, then illustrate that data are produced node and can not be produced enough data, then the data in the buffer area belonging to this variable have all been consumed rear stopping data readback operation; If the data length of this buffer area acquisition last time is greater than the length of request, illustrative system runs makes mistakes, and provides and warns and stop testing playback operation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310399154.8A CN103440200B (en) | 2013-09-05 | 2013-09-05 | A kind of height based on dual operating systems real-time big data quantity test back method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310399154.8A CN103440200B (en) | 2013-09-05 | 2013-09-05 | A kind of height based on dual operating systems real-time big data quantity test back method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103440200A CN103440200A (en) | 2013-12-11 |
| CN103440200B true CN103440200B (en) | 2015-12-23 |
Family
ID=49693891
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310399154.8A Active CN103440200B (en) | 2013-09-05 | 2013-09-05 | A kind of height based on dual operating systems real-time big data quantity test back method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103440200B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110908889A (en) * | 2018-09-17 | 2020-03-24 | 千寻位置网络有限公司 | Automatic testing method and device and control equipment |
| CN109557937B (en) * | 2018-12-07 | 2021-07-09 | 江西洪都航空工业集团有限责任公司 | Multi-cycle data storage and playback method for flight simulator |
| CN110138089B (en) * | 2019-06-06 | 2021-12-28 | 国网北京市电力公司 | Batch preset communication method for distribution network automation master station |
| CN110597725B (en) * | 2019-09-19 | 2023-05-05 | 浙江诺诺网络科技有限公司 | Mysql simulation return method, device and equipment |
| CN113031574B (en) * | 2021-05-27 | 2022-05-24 | 北京全路通信信号研究设计院集团有限公司 | Equipment operation process reproduction method and system |
| CN116737673B (en) * | 2022-09-13 | 2024-03-15 | 荣耀终端有限公司 | Scheduling method, equipment and storage medium of file system in embedded operating system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102141962A (en) * | 2011-04-07 | 2011-08-03 | 北京航空航天大学 | Safety distributed test framework system and test method thereof |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070180369A1 (en) * | 2006-01-31 | 2007-08-02 | Carli Connally | Method and apparatus for automatically formatting data based on a best match test result type |
-
2013
- 2013-09-05 CN CN201310399154.8A patent/CN103440200B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102141962A (en) * | 2011-04-07 | 2011-08-03 | 北京航空航天大学 | Safety distributed test framework system and test method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103440200A (en) | 2013-12-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103440200B (en) | A kind of height based on dual operating systems real-time big data quantity test back method | |
| CN109639521B (en) | Block chain performance testing method, device, equipment and storage medium | |
| CN107995169B (en) | Mobile game server framework system based on micro-service architecture | |
| CN103425538B (en) | Process communication method and system | |
| CN104506589B (en) | A kind of resource migration dispatching method based on super fusion storage | |
| CN103078941B (en) | A kind of method for scheduling task of distributed computing system | |
| CN107809467B (en) | Method for deleting container mirror image data in cloud environment | |
| CN102521014B (en) | Deploying method and deploying device for virtual machine | |
| CN107092541A (en) | A kind of method that utilization virtual machine carries out storage performance test | |
| CN108196958A (en) | Scheduling of resource distribution method, computer system and super fusion architecture system | |
| CN101150485A (en) | A management method for network data transmission of zero copy buffer queue | |
| CN110308984A (en) | It is a kind of for handle geographically distributed data across cluster computing system | |
| Long et al. | A toolkit for modeling and simulating cloud data storage: An extension to cloudsim | |
| WO2024046188A1 (en) | I/o unloading method and system in cloud environment, device, and storage medium | |
| CN101150486A (en) | A management method for receiving network data of zero copy buffer queue | |
| CN103488564B (en) | A kind of distributed real-time testing system multi-channel test data compression and merging method | |
| CN103984240A (en) | Distributed real-time simulation method based on reflective memory network | |
| CN106055590A (en) | Power grid data processing method and system based on big data and graph database | |
| CN102857560A (en) | Multi-service application orientated cloud storage data distribution method | |
| CN115150471B (en) | Data processing method, apparatus, device, storage medium, and program product | |
| CN105677763A (en) | Image quality evaluating system based on Hadoop | |
| CN106453618A (en) | Remote sensing image processing service cloud platform system based on G-Cloud cloud computing | |
| CN116304390B (en) | Time sequence data processing method and device, storage medium and electronic equipment | |
| CN100531070C (en) | Network resource scheduling simulation system | |
| CN108733701A (en) | A kind of query page buffer control method applied to online education |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | 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 |