CN103136217B - A kind of distributed data method for stream processing and system thereof - Google Patents
A kind of distributed data method for stream processing and system thereof Download PDFInfo
- Publication number
- CN103136217B CN103136217B CN201110378247.3A CN201110378247A CN103136217B CN 103136217 B CN103136217 B CN 103136217B CN 201110378247 A CN201110378247 A CN 201110378247A CN 103136217 B CN103136217 B CN 103136217B
- Authority
- CN
- China
- Prior art keywords
- data
- stream
- real
- processing
- time
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 claims abstract description 29
- 230000004044 response Effects 0.000 abstract description 5
- 230000000875 corresponding Effects 0.000 description 9
- 238000010586 diagram Methods 0.000 description 5
- 230000006399 behavior Effects 0.000 description 3
- 238000007405 data analysis Methods 0.000 description 3
- 230000011218 segmentation Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000005540 biological transmission Effects 0.000 description 1
- 238000011030 bottleneck Methods 0.000 description 1
- 230000001351 cycling Effects 0.000 description 1
- 239000004744 fabric Substances 0.000 description 1
- 238000005206 flow analysis Methods 0.000 description 1
- 238000005111 flow chemistry technique Methods 0.000 description 1
- 238000005192 partition Methods 0.000 description 1
- 238000007634 remodeling Methods 0.000 description 1
- 230000004043 responsiveness Effects 0.000 description 1
- 238000000926 separation method Methods 0.000 description 1
- 230000026676 system process Effects 0.000 description 1
- 238000011144 upstream manufacturing Methods 0.000 description 1
Abstract
This application provides a kind of distributed data method for stream processing, described method includes: original data stream is divided into real-time stream and historical data stream；Real-time stream described in parallel processing and described historical data stream, and produce respective result respectively；And produced result is integrated.Present invention also provides a kind of distributed data current processing device, described device includes: data identification module, for original data stream is divided into real-time stream and historical data stream；Parallel processing module, for real-time stream described in parallel processing and described historical data stream, and produces respective result respectively；And Data Integration module, for produced result is integrated.What the application made big data quantity is calculated as possibility in real time, the computing of real-time stream can parallel processing the most in a distributed manner, ensure that big data quantity processes and high real-time simultaneously, improve the response speed of system.
Description
Technical field
The application relates to distributed data processing, particularly relates to a kind of for processing the distributed of big data quantity
Data flow processing method and system thereof.
Background technology
At present, Data Stream Processing becomes the major way of data mining, data analysis.Such as, website day
Will is exactly the data stream of a big data quantity.The most such as, in e-commerce website, ever-increasing commodity are sent out
Cloth information, ever-increasing SMS sends record etc..Such data stream has a characteristic that
(1) data volume is big；(2), in every information, there is the ID (identifier) of feature to be analyzed；(3)
There is time attribute, i.e. timing.
Data-flow analysis usually requires that real-time, quick so that system can be according to particular user current behavior
Make real-time response.Such as, the analysis in real time of daily record can hold the current state of user, nearest visit
Ask behavior, can be effectively improved the precision of recommendation, or the most counter practise fraud.And the most quickly analyze
Data stream, the most in cases where an amount of data is large, meets requirement of real time and is always technical difficult point.
Generally, the ultimate principle of existing distributed data stream processing system is as it is shown in figure 1, initial data
Stream S is assigned to multiple functional module F.Multiple functional modules F process simultaneously, and by after process
Result be all sent to Data Integration module I, Data Integration module I after carrying out integrating and exporting integration
Data.But have following defects that in existing distributed data stream processing system
(1) when processing data stream, when data volume is the biggest when, data process, data are divided
Analysis becomes quite time-consuming.And existing distributed data stream processing system typically uses shared memory module,
That is, between disparate modules, especially between upstream and downstream module, the mode of interaction data is the result of modules A
Being put in storage (data base, file etc.), then module B reads data in storage, is achieved in
Modules A and the data interaction of module B, this pattern can not calculate in real time, can only accomplish quasi real time.
It is to say, when speed becomes bottleneck, and most of existing treatment technology can not meet real-time stream
Growth rate, data delay is bigger so that data analysis can only be carried out by off-line, cause data analysis,
Data mining postpones, it is impossible to make a response the current of user or recent behavior.
(2) for the process of big data quantity, Distributed Parallel Computing has become as trend.And it is existing
Concurrent computational system, is substantially all and is only limitted to the reproducible framework of merit, i.e. system implements parallel calculates
Method be all of computing module be same function, run same program, simply operational data is not
Same part, reaches the purpose of parallel computation with this, therefore cannot realize the most fine-grained parallel, also cannot
Realize the hot plug of modularity and module, and be unfavorable for safeguarding.
Summary of the invention
This application provides a kind of distributed data method for stream processing, described method includes: by initial data
Flow point is slit into real-time stream and historical data stream；Real-time stream described in parallel processing and described history number
According to stream, and produce respective result respectively；And produced result is integrated.
Preferably, in the step processing described real-time stream, described real-time stream is cut by dimension
Divide and carry out parallel processing.
Preferably, the step processing described real-time stream includes: be cut into many by described real-time stream
Individual data block；Concurrently each of the plurality of data block is cut into multiple data cell, then will
The plurality of data cell is sent respectively to multiple different functional module and carries out parallel processing；And will also
The result that row processes collects.
Preferably, in the step processing described historical data stream, described historical data stream is cut by dimension
Divide and carry out parallel processing.
Present invention also provides a kind of distributed data current processing device, described device includes: data identification
Module, for being divided into real-time stream and historical data stream by original data stream；Parallel processing module,
For real-time stream described in parallel processing and described historical data stream, and produce respective process knot respectively
Really；And Data Integration module, for produced result is integrated.
Preferably, described parallel processing module is when processing described real-time stream, to described real time data
Stream is by dimension cutting and carries out parallel processing.
Preferably, process described real-time data processing system and include: transversally cutting module, for by described
Real-time stream is cut into multiple data block；Multiple longitudinal cutting modules, for concurrently by the plurality of
Each of data block is cut into multiple data cell, then the plurality of data cell is sent respectively to
Multiple different functional modules carry out parallel processing；And result summarizing module, for by parallel processing
Result collects.
Preferably, described parallel processing module is when processing described historical data stream, to described historical data
Stream is by dimension cutting and carries out parallel processing.
Distributed data method for stream processing according to the application, by sequence on time and by dimension to data stream
Carry out repeated segmentation and cutting, i.e. utilize timing, use multiple structure, data are processed at times,
Use new distributed structure/architecture, utilize different dimensions, flow of information is carried out longitudinal cutting.Make big data
That measures is calculated as possibility in real time.The computing of real-time stream can be located the most parallel
Reason, ensure that big data quantity processes and high real-time simultaneously, improves the response speed of system.
Accompanying drawing explanation
Below with reference to appended accompanying drawing, embodiments herein is described, wherein:
Fig. 1 illustrates the schematic diagram of the distributed data stream processing system of prior art；
Fig. 2 illustrates an enforcement of the big data quantity distributed data stream processing system of the application
The schematic diagram of example；
Fig. 3 illustrates corresponding with the big data quantity distributed data stream processing system in Fig. 2
The flow chart of the big data quantity distributed data method for stream processing of the application；
Fig. 4 illustrates the schematic diagram of an embodiment of the real time processing system 30 in Fig. 2；With
And
Fig. 5 illustrates the real-time place of the application corresponding with the real time processing system 30 in Fig. 4
The flow chart of reason method.
Detailed description of the invention
Above-mentioned spirit and the essence of the application is described in detail below in conjunction with Fig. 2-Fig. 5.
Although describing the embodiment party of the system and method for the application below as a example by web log file data stream
Formula, it will be understood that the application may also be used for processing personalized recommendation, the most anti-cheating, commodity are issued,
The data stream of the systems such as SMS transmission, scientific algorithm.
As a example by web log file data stream, Fig. 2 illustrates the distributed number of big data quantity of the application
Schematic diagram according to an embodiment of stream processing system.
Big data quantity distributed data stream processing system in Fig. 2 includes: data identification module 10；30
Data handling system 20 before it；Real-time data processing system 30；Data handling system 40 within 30 days；
And Data Integration module 50.It is appreciated that these modules can be by a computer or similar having
The network or the one of such equipment that calculating or the equipment of disposal ability or the such equipment of multiple stage are formed
Fractional hardware or software realize.
Fig. 3 illustrates corresponding with the big data quantity distributed data stream processing system in Fig. 2
The flow chart of the big data quantity distributed data method for stream processing of the application.Come below in conjunction with Fig. 2 and Fig. 3
One embodiment of the application is described.
In step S100, obtain original data stream 100.
In step S101, after original data stream 100 is obtained by data identification module 10, data identification
Module 10 identify the data in original data stream 100 be real time data or the data within 30 days,
Or the data before 30 days, thus before original data stream 100 sequence on time is divided into 30 days
Data stream 200, real-time stream 300 and data stream 400 within 30 days.Data before 30 days
Stream 200 is sent to 30 days data handling systems 20 in the past, and real-time stream 300 is sent to reality
Time data handling system 30, and 30 days within data stream 400 be sent to data within 30 days
Processing system 40.
In step S102, before before 30 days, data handling system 20 is carried out 30 days, data process,
Result is sent to Data Integration module 50.In step S103, real-time data processing system 30
Carry out real time data processing, result is sent to Data Integration module 50.In step S104,
Within within 30 days, data handling system 40 is carried out 30 days, data process, and result are sent to
Data Integration module 50.Step S102, step S103 and step S104 executed in parallel.
In step S105, the result received is integrated by Data Integration module 50, and defeated
Go out the data after integrating.
Although being appreciated that here, original data stream 100 is divided into by data identification module 10
Data stream 200 before 30 days, real-time stream 300 and within 30 days data stream 400 such
The different piece distinguished by three time dimensions, those skilled in the art can be according to practical situation with it
Its time dimension splits original data stream 100.Such as, original data stream 100 is divided into less
Or (correspondingly, big data quantity distributed data stream processing system comprises less or more more time period
Many data handling systems), or use the time dimension being different from 30 days, or according to the actual requirements
Define the time range by being counted as " in real time ".
By above embodiment it will be seen that the big data quantity distributed data method for stream processing of the application
Be essentially divided into sequence segmentation on time, data process, the such three phases of Data Integration.
The stage is split, owing to system journal is to add, therefore, first by data in the moment in sequence on time
Real-time stream 300 is distributed to real time processing system 30 by identification module 10；For historical data (example
Such as data streams 200 and data stream 400 within 30 days before 30 days), owing to they have been stored as
File, thus be sent to history file processing system (before such as 30 days data handling systems 20 and
Data handling system 40 within 30 days).
At data processing stage, history processing system and real time processing system process different periods concurrently
Data.
Result after Data Integration stage, the parallel data processing of different periods is sent to number
According to integrating module 50, after these results are integrated, it is possible to output, provide service with external.
In the present embodiment, system and data are split by sequence on time, are very beneficial for process and have
The data stream of the big data quantity of timing, this is the basis that the application processes mass data.
Every information of imagination data stream has timestamp, then from the data started most data till now
(still in growth), it is simply that the data stream of full dose.If certain time point is set to separation, then
(whole) data of this full dose can be divided into historical data and real time data.For full dose
Data stream, we can analyze, the historical data before certain period, before certain time point
Exist.Within such as one day, front data need not calculate in real time, it is possible to calculated off line, it is only necessary to
The result of its result of calculation He other modules (such as real-time processing module) is integrated.
Process respectively by historical data and real-time calculating, to historical data calculated off line, can greatly subtract
The light pressure calculated in real time.Real time data is enable to be calculated quickly.Meanwhile, historical data can obtain
To finer calculating.
The application sequence partition data on time, makes the data of different periods process and can carry out parallel, thus
Ensure that the high responsiveness energy of real time data.
In order to improve the performance of real-time data processing system further, the application also proposed the letter of data
By dimension, (in this application, " dimension " word is used for distinguishing different attribute interest statement unit (i.e. data block)
Or the data of the data of type, i.e. different dimensions are processed by dissimilar functional module) cutting further
To each functional module (the most different types of functional module).Below will be with real-time data processing system 30
As a example by illustrate.
Fig. 4 illustrates the schematic diagram of an embodiment of the real time processing system 30 in Fig. 2.
As shown in Figure 4, real time processing system 30 includes: one (the most in this application, " the most laterally "
One word be for only for ease of identify this level cutting, rather than the concept on direction) cutting module 400；
Multiple (N number of) longitudinally (in this application, " longitudinally " word is for only for ease of this level of mark
Cutting, rather than the concept on direction) cutting module 500；Multiple (N number of) functional module group 600,
The most each functional module group 600 comprises multiple (M) functional module；And result summarizing module 700.
Fig. 5 illustrates the real-time place of the application corresponding with the real time processing system 30 in Fig. 4
The flow chart of reason method.One of real time processing system of the application is described below in conjunction with Fig. 4 and Fig. 5
Embodiment.
In step S200, obtain real-time stream 300.
In step S201, the real-time stream 300 of acquisition is cut into many numbers by transversally cutting module 400
According to block (1,2,3...N...), (the most so-called transversally cutting of cutting in the step for of), and will
The data block of institute's cutting is sent respectively to multiple (N number of) longitudinal cutting module 500.As shown in Figure 4,
1st data block is sent to the 1st longitudinal cutting module 500, the 2nd data block is sent to the 2nd
Individual longitudinal cutting module 500, by that analogy, is sent to n-th longitudinal direction dividing die by n-th data block
Block 500.It is appreciated that, it is contemplated that although data stream is unlimited, but flowing, and multiple (N
Individual) each in longitudinal cutting module 500 can be reused after having processed a data block,
Therefore, arrange can be depending on the uninterrupted of data stream for the quantity of longitudinal cutting module 500.
In step S202, received data block is cut into multiple by cutting module longitudinally in each 500
(depending on practical situation, can up to M) (cutting in the step for of be the most so-called by dimension for data cell
Longitudinal cutting), and multiple data cells of institute's cutting are sent respectively in a functional module group 600
Multiple (corresponding to the quantity of data cell, up to M) different functional module.
As shown in Figure 4, data block 1 is cut into M data sheet by the 1st longitudinal cutting module 500
Unit, and the 1st data cell is sent the 1st functional module to the 1st functional module group 600,
2nd data cell is sent the 2nd functional module to the 1st functional module group 600, with this type of
Push away, m-th data cell is sent the m-th functional module to the 1st functional module group 600.
By that analogy, if the data traffic of real-time stream 300 is sufficiently large, the 2nd longitudinal dividing die
Data block 2 is cut into M data cell by block 500, and sends the 1st data cell to the 2nd
1st functional module of functional module group 600, sends the 2nd data cell to the 2nd function mould
2nd functional module of block group 600, by that analogy, sends m-th data cell to the 2nd merit
The m-th functional module of energy module group 600.
By that analogy, if the data traffic of real-time stream 300 is sufficiently large, more number can be there is
According to block, longitudinal cutting module 500, functional module group 600 and functional module.It is appreciated that longitudinal direction
The quantity of the functional module in cutting module 500, functional module group 600 and functional module group 600 sets
Put and can depend on the needs respectively.
Step S202 and step S203 executed in parallel.
In step S203, received data unit is processed by each functional module, and by after process
Result be sent to result summarizing module 700.
In step S204, the result received is collected by result summarizing module 700, and output collects
After data.
By the description of the present embodiment, it can be seen that first, real-time stream can be divided by transversally cutting
Being fitted on each processor (such as longitudinal cutting module 500), the function of each processor is the same.
These processor parallel processings, drastically increase processing speed.
Then, data block by dimension longitudinal direction cutting, is i.e. extracted from data block by longitudinal cutting module 500
Going out the data cell of different dimensions, then the data cell of respective dimensions is sent to corresponding function treatment
Module (i.e. functional module), by each functional processing module parallel processing.
As a example by web log file data stream is as original data stream, web log file data stream is first by transversally cutting
Multiple log information data block, each log information data block is become to be assigned to a corresponding longitudinal cutting
In module 500.Then, corresponding log information data block is indulged by each longitudinal cutting module 500 by dimension
To cutting, such as, from log information data block, extract merchandise news deliver to commodity processing unit, extract
Key word information delivers to key word processing unit.So, each information unit is broken down into more particulate
The element of degree, is distributed to each functional unit, parallel processing.Such as, as processing real-time website daily record
The functional unit of data stream, such as, merchandise news parsing module resolves merchandise news, access path module
Resolve access path, modules parallel processing.Then, user and merchandise news are sent to recommendation function
Module, user and access path information are sent to anti-module of practising fraud, and modules is also parallel processing.
Finally, the result that each functional module processes, all it is sent to integrator (such as result summarizing module
700, or the most also include Data Integration module 50) on, integrator result is integrated (collecting)
Process.
As a example by real-time data processing system, describe the application above data are carried out cutting.It is appreciated that
For the processing system of historical data, similar framework can be used.Except for the difference that, due to historical data
Process the mode using cycling service, it is possible to use the cluster distributed arithmetic system of low cost.
Pass through above description, it can be seen that the application is not as existing distributed traffic processes
System uses shared memory module like that, but by sequence on time and carries out many as dimension to data stream
Secondary segmentation and cutting, i.e. utilize timing, uses multiple structure, processes data at times, uses
New distributed structure/architecture, utilizes different dimensions, and flow of information carries out longitudinal cutting, rather than as existing
Concurrent computational system is only limitted to the reproducible framework of merit like that, say, that the present invention realizes parallel computation
Method be not all of computing module be same function, run same program, simply operand
According to different piece.Therefore the present invention be capable of the most fine-grained parallel, it is also possible to realize modularity and
The hot plug of module, and be conducive to safeguarding.
The invention enables big data quantity is calculated as possibility in real time.The computing of real-time stream can be
Limits ground parallel processing in a distributed manner, ensure that big data quantity processes and high real-time simultaneously, improves
The response speed of system.
Big data quantity distributed data method for stream processing according to the application can be by having calculation process
The single or multiple processing equipments of ability, such as single or multiple computers, run computer and can perform to refer to
Order realizes.Big data quantity distributed data stream processing system according to the application can be single or many
Individual processing equipment, if single or multiple computers, modules therein or unit can be this process
Equipment has the apparatus assembly of corresponding function when running computer executable instructions.According to the application one
Individual embodiment, it is possible to use the language such as JAVA, SQL are under the systems such as linux, Windows
Realize above-mentioned big data quantity distributed data method for stream processing and system thereof.
Although with reference to exemplary embodiment describing the application, it is to be understood that, term used is to say
Bright and exemplary and nonrestrictive term.Owing to the application can be embodied as in a variety of forms
Without departing from spirit or the essence of invention, it should therefore be appreciated that above-described embodiment is not limited to any aforesaid
Details, and should explain widely in the spirit and scope that appended claims are limited, therefore fall into
Whole changes and remodeling in claim or its equivalent scope all should be appended claims and contained.
Claims (6)
1. a distributed data method for stream processing, described method includes:
Original data stream is divided into real-time stream and historical data stream；
Real-time stream described in parallel processing and described historical data stream, and produce respective process knot respectively
Really；And
Produced result is integrated；Wherein, in the step processing described historical data stream
Calculated off line is used in rapid,
In the step processing described real-time stream, described real-time stream by dimension cutting and is carried out
Parallel processing, the most fine-grained parallel with realization.
Method the most according to claim 1, wherein, processes the step bag of described real-time stream
Include:
Described real-time stream is cut into multiple data block；
Concurrently each of the plurality of data block is cut into multiple data cell, then by described many
Individual data cell is sent respectively to multiple different functional module and carries out parallel processing；And
The result of parallel processing is collected.
Method the most according to claim 1, wherein,
In the step processing described historical data stream, described historical data stream by dimension cutting and is carried out
Parallel processing.
4. a distributed data current processing device, described device includes:
Data identification module, for being divided into real-time stream and historical data stream by original data stream；
Parallel processing module, for real-time stream described in parallel processing and described historical data stream,
And produce respective result respectively；And
Data Integration module, for integrating produced result；Wherein,
Described parallel processing module uses calculated off line when processing described historical data stream,
Described parallel processing module is when processing described real-time stream, to described real-time stream by dimension
Cutting also carries out parallel processing, the most fine-grained parallel with realization.
Device the most according to claim 4, wherein, processes described real-time data processing system bag
Include:
Transversally cutting module, for being cut into multiple data block by described real-time stream；
Multiple longitudinal cutting modules, for being cut into multiple by each of the plurality of data block concurrently
Data cell, is then sent respectively to the plurality of data cell multiple different functional module and carries out also
Row processes；And
Result summarizing module, for collecting the result of parallel processing.
Device the most according to claim 4, wherein,
Described parallel processing module is when processing described historical data stream, to described historical data stream by dimension
Cutting also carries out parallel processing.
Priority Applications (7)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110378247.3A CN103136217B (en) | 2011-11-24 | A kind of distributed data method for stream processing and system thereof | |
| TW101107358A TW201322022A (en) | 2011-11-24 | 2012-03-05 | Distributed data stream processing method |
| US13/681,271 US9250963B2 (en) | 2011-11-24 | 2012-11-19 | Distributed data stream processing method and system |
| EP12806741.0A EP2783293A4 (en) | 2011-11-24 | 2012-11-20 | Distributed data stream processing method and system |
| PCT/US2012/066113 WO2013078231A1 (en) | 2011-11-24 | 2012-11-20 | Distributed data stream processing method and system |
| JP2014537381A JP6030144B2 (en) | 2011-11-24 | 2012-11-20 | Distributed data stream processing method and system |
| US14/977,484 US9727613B2 (en) | 2011-11-24 | 2015-12-21 | Distributed data stream processing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201110378247.3A CN103136217B (en) | 2011-11-24 | A kind of distributed data method for stream processing and system thereof |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103136217A CN103136217A (en) | 2013-06-05 |
| CN103136217B true CN103136217B (en) | 2016-12-14 |
Family
ID=
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6850947B1 (en) * | 2000-08-10 | 2005-02-01 | Informatica Corporation | Method and apparatus with data partitioning and parallel processing for transporting data for data warehousing applications |
| CN101276364A (en) * | 2007-03-30 | 2008-10-01 | 阿里巴巴公司 | Method, system and apparatus for combining distributed computational data |
| CN101694667A (en) * | 2009-10-19 | 2010-04-14 | 东北电力大学 | Distributed data digging method for intelligent electrical network mass data flow |
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6850947B1 (en) * | 2000-08-10 | 2005-02-01 | Informatica Corporation | Method and apparatus with data partitioning and parallel processing for transporting data for data warehousing applications |
| CN101276364A (en) * | 2007-03-30 | 2008-10-01 | 阿里巴巴公司 | Method, system and apparatus for combining distributed computational data |
| CN101694667A (en) * | 2009-10-19 | 2010-04-14 | 东北电力大学 | Distributed data digging method for intelligent electrical network mass data flow |
Non-Patent Citations (2)
| Title |
|---|
| MESHJOIN*：实时数据仓库环境下的数据流更新算法;林子雨 等;《Journal of Frontiers of Computer Science and Technology 计算机科学与探索》;20101015(第10期);927-939 * |
| 并行ETL过程的研究与实现;许力 等;《Computer Engineering and Applications 计算机工程与应用》;20090501(第13期);170-172 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US9727613B2 (en) | Distributed data stream processing method and system | |
| CN104090886B (en) | The method that structure user draws a portrait in real time and device | |
| CN110058977B (en) | Monitoring index abnormity detection method, device and equipment based on stream processing | |
| KR101559206B1 (en) | Method of processing log data, and system operating the same | |
| CN107577805A (en) | A kind of business service system towards the analysis of daily record big data | |
| US20150032759A1 (en) | System and method for analyzing result of clustering massive data | |
| CN106506266B (en) | Network flow analysis method based on GPU, Hadoop/Spark mixing Computational frame | |
| Roschke et al. | A flexible and efficient alert correlation platform for distributed ids | |
| CN105306463A (en) | Modbus TCP intrusion detection method based on support vector machine | |
| CN109981416A (en) | A kind of block chain method for testing performance and system | |
| CN110472154A (en) | A kind of resource supplying method, apparatus, electronic equipment and readable storage medium storing program for executing | |
| CN106815254A (en) | A kind of data processing method and device | |
| CN113609201A (en) | Service data processing method and system | |
| CN112182043A (en) | Log data query method, device, equipment and storage medium | |
| CN106557483B (en) | Data processing method, data query method, data processing equipment and data query equipment | |
| CN109542657A (en) | The processing method and server of system exception | |
| CN104487942A (en) | Event correlation | |
| CN103136217B (en) | A kind of distributed data method for stream processing and system thereof | |
| CN111768219B (en) | Advertisement crowd experiment method, device and storage medium | |
| CN115713395A (en) | Flink-based user wind control management method, device and equipment | |
| CN108363761A (en) | Hadoop awr automatic loads analyze information bank, analysis method and storage medium | |
| Banik et al. | On the finite buffer queue with renewal input and batch Markovian service process: GI/BMSP/1/N | |
| CN114490262A (en) | Database monitoring method, device, equipment and storage medium | |
| Sahin et al. | Streaming event detection in microblogs: Balancing accuracy and performance | |
| Peng et al. | On a BMAP/G/1 G-queue with setup times and multiple vacations |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20211103 Address after: Room 201, floor 2, building 15, No. 1999, Yizhou Avenue, hi tech Zone, Chengdu, Sichuan Patentee after: Alibaba (Chengdu) Software Technology Co.,Ltd. Address before: A four-storey 847 mailbox in Grand Cayman Capital Building, Cayman Islands Patentee before: ALIBABA GROUP HOLDING Ltd. |