REAL-TIME HIGH-SPEED STREAM DATA PROCESSING METHOD BASED ON DISTRIBUTED ENVIRONMENT
The present invention refers to relates to distributed computing techniques, high speed processing is provided based on distributed environment using the same method and embodiment stream data between a DBMS (Data Base Management System) are disclosed. While specific event stream data processing such as invitation dosage have the meanings in store is queried to same when analyzing existing DBMS, a performance degradation and extremely inefficient door number management is performed with each other. In addition, existing stream data processing of the comb-shaped stream data processing center technology that are limited, various kinds of stream data processing techniques for low level shifted disclosed. In addition, a stream of data processing is based on existing JVM (Java Virtual Machine) embodiment is suited to perform the analysis, due to the problem of structural door number JVM dependency and memory management (JVM heap: Garbage Collection performed), the disadvantages of lower CPU office environment setting and dependent library has a door number JVM and packaging. In order to solve the present invention refers to said door number point such as battery is provided, the purpose of the invention is, without the problem of JVM structural door number, a stream data between various embodiment to generate mass analyzed data analysis method using the same and a number [...] DBMS is streams can be recycled. In the embodiment according to one of the present invention for achieving said purpose, the stream data analysis method, filter module, determining whether the data stream; processor, analysis determined data for analysing generating task; and said task, analysis determined data analyzing; without using a tool. And, said filter module, capable of changing setting environment can be. In addition, said filter module, can be further conflict. And, said that the stream data, shaping stream data, the stream data and non-stabilizing system comprising at least one stream data can. In addition, said that the stream data, be a mass data generated events. And, said analysis step, based on distributed environment can be performed. In addition, said analysis step, between embodiment can be performed. On the other hand, in the embodiment according to other of the present invention, the DBMS, stream data is stored DB; and stream data or the analysis of, for analysing determined data analysis task analyzing data processor generates analysis determined; it contains. As the device, according to of the present invention in the embodiment, the problem of JVM without structural door number, mass stream data can be analyzed to generate various embodiment resulting in between. The present invention also applicable example Figure 1 shows a side stream data analyzer, Figure 2 shows a also, through the probe and analyzer stream data analysis also 1 stream data shown in indicating the card information, Figure 3 shows a also, stream data analyzer 2 also shown in the block indicating the structure of DBMS can be implemented, and, Figure 4 shows a big data analysis method according to an embodiment of the present invention ball number to the elucidation of flow which also are disclosed. The present invention hereinafter with reference to the drawing in more detailed as follows. The present invention also applicable example Figure 1 shows a side stream data analyzer are disclosed. The present invention stream data analyzer applicable, distributed environment embodiment is based on a high speed stream data operator for processing are disclosed. The present invention stream data analyzer applicable form event in a variety of data sources, having large amounts of continuity between high speed skyrocketing [...] stream data management can be embodiment and analyzing, "Squall" referred to him. The present invention applicable stream data analyzer, as shown in fig. 1, query (SQL parser, Query Optimizer) for processing modules, for processing modules (CLI, JDBC) command, etc. for managing cache module (Cache Manager). In addition, the present invention stream data analyzer applicable, distributed environment embodiment is based on a high speed stream data analysis and for managing data processing architecture. The stream data, shaping stream data (statistical analysis can be used to form dependable: regional demographics), semi-stream data (statistical analysis directly using positive enough not number data: newspaper) and non-stream data etc. (Raw data form: moving picture, photograph, SNS full text). Figure 2 shows a also, through the probe and analyzer stream data analysis is also 1 stream data shown in indicating the process are disclosed. Shard structure within the adapted, according to an embodiment of the present invention stream data analyzer based distributed environment are disclosed. In addition, stream data analyzer according to an embodiment of the present invention is embodiment processing stream data to direct the other. As shown in fig. 2, stream data analyzer, StreamContext module (110), Stream module (120), StreamShard module (130), StreamProcessor module (140) and Task module (150) comprises. StreamContext module (110) is shaping stream data, semi-stream data, non-stream data or a combination of these such as stream data module receives stream data are disclosed. Stream module (120) is StreamContext module (110) via a stream data StreamShard module (130) 620, StreamShard module (130) of the transferred data is stored to a distributed DB Shard stream. In addition, Stream module (120) data analysis request stream data filtered from the other. Cells such as, Stream module (120) equipped only augmented query is stream data filtering substrate. Query conditions, cells such as, data generating period, data generating sensor ID will other. Thus, Stream module (120) that can be the determining whether analysis of the stream data. Stream module (120) and capable of changing the setting environment, process from extended (additional generating). Distributed environment or stream data type is set according to the environment as well as change and extensible, abridgment pivotably. StreamProcessor module (140) is Stream module (120) for analyzing data are analyzed by a decided Task module (150) is defined. Task module (150) deciding to the analysis data StreamShard module (130) may be obtained from from analyzing, stream data analysis results StreamProcessor module (140) having traversed substrate. Figure 3 shows a also, stream data analyzer can be implementing 2 also shown in block indicating the structure of DBMS are disclosed. According to an embodiment of the present invention DBMS is, as shown in fig. 3, NIC (Network Interface Card) (210), processor (220) and DB (230) comprises. NIC (210) comprises an outer system, to make access to the external network communication is communication means are disclosed. NIC (210) through the inflow stream data are disclosed. In Figure 2 the aforementioned StreamContext module (110) is NIC (210) corresponding to each other. DB (230) with big data is stream data stored large capacity disk storage are disclosed. DB (230) includes a plurality of [...] and composed, in Figure 2 the aforementioned StreamShard module (130) is DB (230) corresponding to each other. Processor (220) the streams or the analysis, determined data analysis has a plurality of hierarchies. In Figure 2 the aforementioned StreamProcessor module (140) and Task module (150) the processor (220) corresponding to each other. By big data analysis method number 4 in the auditory canal 3 also shown in DBMS being disclosed. Figure 4 shows a big data analysis method to the elucidation of the surface thereof on which ball number according to an embodiment of the present invention also are disclosed. As shown in fig. 4, stream data analysis request if (S310 a-Y), processor (220) is subject only filtering substrate (S320) query condition with reference to stream data of an analyte. Then, processor (220) is filtered data for analyzing generate tasks (S330), tasks (S340) produce a result embodiment stores the data between establishment of a branch. On the other hand, in the embodiment according to device a method a computer program to carry out tasks of the computer-readable recording medium of the present invention is applicable to even the engine technical idea are disclosed. In addition, in the embodiment according to of the present invention idea is to various computer-readable recording medium recording computer-readable code embodied in the form of disapproval. Computer-readable recording medium readable by a computer and data from the capable of pivotably even data storage device. For example, computer-readable recording medium includes a ROM, RAM, CD BD-ROM, magnetic tape, floppy disk, optical disk, hard disk drive, is red are installed are disclosed. In addition, computer-readable recording medium stored on a computer-readable computer code or program transmitted over a network coupled between the disapproval. In addition, or more of the present invention preferred embodiment shown in and described with respect to but, in the embodiment described above the present invention refers to the use of a specific defined correspondingly, in the claims claim the subject matter of invention made without deviating from the invention belongs in the art 155.520 thereby enabling as well as various modified embodiment, these modified embodiment of the present invention are technical idea must not be understood to or separately from the outlook will. 110: StreamContext module 120: Stream module 130: StreamShard module 140: StreamProcessor module 150: Task module A real-time high-speed stream data processing method based on a distributed environment is provided. According to an embodiment of the present invention, a method for analyzing stream data comprises the following steps. A filter module determines an analysis state of stream data. A processor generates a task to analyze data determined to be analyzed. And the task analyzes the data determined to be analyzed. Accordingly, a large quantity of various stream data can be analyzed in real time without a structural problem that a java virtual machine (JVM) has. COPYRIGHT KIPO 2017 Filter module, determining whether the data stream; processor, analysis determined data for analysing generating task; said task, analysis determined data analyzing; characterized in including a stream data analysis method. According to Claim 1, said filter module, capable of changing setting environment characterized stream data analysis method. According to Claim 1, said filter module, it is possible to generating additional characterized stream data analysis method. According to Claim 1, said that the stream data, shaping stream data, the stream data and non-stream data including at least one stabilizing system characterized stream data analysis method. According to Claim 1, said that the stream data, a stream data analysis method characterized in that a large amount of data generated events. According to Claim 1, said analysis step, carried out based on distributed environment characterized stream data analysis method. According to Claim 1, said analysis step, between embodiment characterized stream data analysis method is performed. Stream data is stored DB; or the analysis of stream data, analyzing data determined analysis for analyzing data determined analyzed a preview pane processor; including a characterized the DBMS.
