KR20160063182A - Hadoop-based system for providing distributed multimedia streaming service and method thereof - Google Patents

Abstract

Disclosed are a system for providing a distributed multimedia streaming service and a method for providing a streaming service which can select a streaming server having an optimal condition for a streaming service. According to an embodiment of the present invention, the method for providing a content data streaming service comprises: a step of selecting an optimal streaming server among streaming servers based on at least one among a use rate of a physical resource and a streaming transmission processing amount; and a step of using the optimal streaming server to stream content data.

Description

TECHNICAL FIELD [0001] The present invention relates to a distributed multimedia streaming service providing system and a streaming service providing method,

The following embodiments are directed to a system for providing a distributed multimedia streaming service and a method for providing a streaming service.

The emergence of high-performance heterogeneous smart devices and the proliferation of social networking services (SNS), such as Facebook and Twitter, have resulted in a large number of social media being shared in wired and wireless environments, Content. Recently, the most noteworthy and influential new technologies include cloud-based media streaming, transcoding, and content distribution that provide multi-media services that ensure quality of service (QoS).

Previously, traditional distributed and cluster-based computing approaches were used for media transcoding and streaming processing. However, it is difficult to provide a variety of media services by integrating multi-video traffic and mobile services using a traditional approach.

There are also various approaches for efficiently distributing streaming tasks in a distributed system environment when a large number of requests for streaming services are generated by users. In particular, there are two streaming task distribution methods based on two algorithms. The two algorithms are Round Robin (RR) and Least Connection (LC). However, systems with RR and LC do not consider physical resource usage and streaming network traffic. As a result, they are limited and put a heavy burden on streaming servers and the current Internet infrastructure.

Embodiments can provide a technique of selecting a streaming server having an optimum condition for a streaming service by considering factors of streaming network traffic and usage rate of physical resources.

Accordingly, embodiments can provide a rich and rich media streaming service through a streaming server having optimal conditions in a cloud computing environment having unpredictable network traffic.

That is, embodiments can improve the streaming task distribution capacity.

According to an embodiment of the present invention, there is provided a content data streaming method comprising: selecting an optimal streaming server among streaming servers based on at least one of a usage rate of physical resources and a streaming transmission throughput; .

The selecting may include calculating a usage rate of physical resources of each of the streaming servers, and calculating a streaming transmission throughput of each of the streaming servers.

The selecting may include selecting a streaming server having the lowest usage rate of the physical resources among the streaming servers as the optimal streaming server.

The selecting may further include selecting a streaming server having the lowest streaming transmission throughput among the streaming servers as the optimal streaming server when the number of streaming servers having the lowest usage rate of the physical resources is two or more .

The physical resources may include a CPU and a RAM.

The step of calculating the usage rate of the physical resources may include calculating the usage rate of the physical resources based on the CPU usage rate and the RAM usage rate.

The CPU utilization can be calculated by the lux command mpstat.

The RAM utilization rate can be calculated by a free command.

The step of calculating the streaming transmission throughput may include calculating the streaming transmission throughput based on the current streaming transmission throughput and the previous streaming transmission throughput.

The current streaming transmission throughput may be the total throughput since the start of the streaming service.

The previous streaming transmission throughput may be one second before the point at which the current streaming transmission throughput is measured.

The system for providing a distributed multimedia streaming service according to an exemplary embodiment includes a streaming module including streaming servers and a content data management unit for selecting an optimal streaming server among the streaming servers based on at least one of a usage rate of the physical resources and a throughput of streaming transmission And the optimal streaming server may stream the content data.

The content data management server may calculate a usage rate of physical resources of each of the streaming servers and calculate a streaming transmission throughput of each of the streaming servers.

The content data management server may select a streaming server having the lowest usage rate of the physical resources among the streaming servers as the optimal streaming server.

The content data management server may select a streaming server having the lowest streaming transmission throughput among the streaming servers as the optimal streaming server when there are two or more streaming servers having the lowest usage rate of physical resources.

The physical resources may include a CPU and a RAM.

The content data management server may calculate the usage rate of the physical resource based on the CPU usage rate and the RAM usage rate.

The CPU utilization can be calculated by the lux command mpstat.

The RAM utilization rate can be calculated by a free command.

The content data management server may calculate the streaming transmission throughput based on the current streaming transmission throughput and the previous streaming transmission throughput.

The current streaming transmission throughput may be the total throughput since the start of the streaming service.

The previous streaming transmission throughput may be one second before the point at which the current streaming transmission throughput is measured.

Figure 1 shows a schematic diagram of a redirection-based hierarchical web server.
Figure 2 shows a redirection mechanism when an HTTP request is sent to a redirection server.
3 illustrates a distributed multimedia streaming service providing system according to an exemplary embodiment.
FIG. 4 shows a schematic structure of a system for providing a distributed multimedia streaming service according to an embodiment.
FIG. 5 shows an SRC algorithm used by the management server shown in FIG. 4 for streaming work distribution.
FIG. 6 illustrates a distributed multimedia streaming service providing system according to an example embodiment.

It is to be understood that the specific structural or functional descriptions of embodiments of the present invention disclosed herein are presented for the purpose of describing embodiments only in accordance with the concepts of the present invention, May be embodied in various forms and are not limited to the embodiments described herein.

Embodiments in accordance with the concepts of the present invention are capable of various modifications and may take various forms, so that the embodiments are illustrated in the drawings and described in detail herein. It should be understood, however, that it is not intended to limit the embodiments according to the concepts of the present invention to the particular forms disclosed, but includes all modifications, equivalents, or alternatives falling within the spirit and scope of the invention.

The terms first, second, or the like may be used to describe various elements, but the elements should not be limited by the terms. The terms may be named for the purpose of distinguishing one element from another, for example without departing from the scope of the right according to the concept of the present invention, the first element being referred to as the second element, Similarly, the second component may also be referred to as the first component.

It is to be understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, . On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between. Expressions that describe the relationship between components, such as "between" and "between" or "neighboring to" and "directly adjacent to" should be interpreted as well.

The terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, the terms "comprises ", or" having ", and the like, are used to specify one or more of the features, numbers, steps, operations, elements, But do not preclude the presence or addition of steps, operations, elements, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the meaning of the context in the relevant art and, unless explicitly defined herein, are to be interpreted as ideal or overly formal Do not.

Hereinafter, embodiments will be described in detail with reference to the accompanying drawings.

Figure 1 shows a structure of a redirect-based hierarchical web server, and Figure 2 shows a redirection mechanism when an HTTP request is sent to a redirect server.

Referring to Figures 1 and 2, Figure 1 may be a redirect-based hierarchical web server architecture that facilitates load distribution for HTTP requests.

The categorized content may be stored in distributed servers. A hierarchical web server architecture may include two component levels, for example, redirect servers and normal HTTP servers. Each HTTP server can be categorized and store some of the content data available on the site. The RR DNS (domain name service) may select a redirection server to provide a requested service using the RR algorithm when users request specific content for their system. The selected redirection server may then distribute the task of associating the user request with an HTTP server containing the requested content.

Each file can contain a unique base URL, such as a domain name. You can use the default URL to connect to the redirect server that maps the base URL to the target URL. The target URL may be returned to the user with a redirection message. Next, the user uses a new URL to connect to the HTTP server to obtain specific content.

3 illustrates a distributed multimedia streaming service providing system according to an exemplary embodiment.

Referring to FIG. 3, the distributed multimedia streaming sub providing system 300 may be a cloud-based distributed multimedia streaming service (CloudDMSS) providing system.

The distributed multimedia streaming service providing system 300 may receive the content data, transcode the received content data in a predetermined format, and transmit the content data transcoded to the terminal device 310. [

That is to say, the user of the terminal device 310 can receive the streaming service for the content data transcoded in the format set through the distributed multimedia streaming service providing system 300. The user of the terminal device 310 can select the content data to be provided with the streaming service and can set the format of the content data to be transcoded. The streaming service provided by the distributed multimedia streaming service providing system 300 to the terminal device 310 may be a multi-screen service.

The content data may be media content such as, for example, video content such as a movie, music video or animation, or audio content.

The terminal device 310 may be any device capable of receiving streaming service of content data through the distributed multimedia streaming service providing system 300 such as a personal computer, a notebook, a tablet PC, and a smart phone smartphone) and the like.

The distributed multimedia streaming service providing system 300 can provide a streaming service of content data to one or more terminal devices. For example, the terminal device 310 may be one or more heterogeneous terminal devices.

The content data can be distributed and stored in the transcoding Hadoop cluster.

A user of the content data or an administrator of the distributed multimedia streaming service providing system 300 may store the content data in a transcoding Hadoop cluster for transcoding the content data for sharing the content data with other users. If the content data is uploaded to the transcoding Hadoop cluster, the content data may be transcoded into a standard format. The content data transcoded into the standard format can be streamed to the terminal device 310. [ The standard format may be MPEG4, for example. Transcoding can be performed utilizing a mapreduce framework. Transcoding is performed using the MapReduce framework, thereby reducing the time required for transcoding.

The streaming Hadoop cluster may include one or more streaming servers. The transcoded content data may be transmitted to the terminal device 310 through streaming with quality of service (QoS) guaranteed by one or more streaming servers.

In order to prevent the occurrence of content delay and traffic bottlenecks, a streaming job distribution algorithm that balances and distributes the load of the streaming server may be used for streaming content data.

The structure and function of the distributed multimedia streaming service providing system 300 will be described in detail with reference to FIG. 4 to be described later.

The technical contents described with reference to FIG. 1 and FIG. 2 may be applied as they are, so a detailed description will be omitted below.

FIG. 4 shows a schematic structure of a system for providing a distributed multimedia streaming service according to an embodiment.

In FIG. 4, the solid lines indicate the components of the streaming service system 300, and the dotted lines indicate the function of the configuration including the content.

As described above with reference to FIG. 3, the distributed multimedia streaming service providing system 300 may be designed to operate on a Hadoop cluster.

Referring to FIG. 4, a distributed multimedia streaming service providing system 300 may include a content data management server 410, a transcoding module 420, and a streaming module 430.

For example, the content data management server 410 may be a cloud multimedia management module (CMMM) as a content data management module. The transcoding module 420 may be a Hadoop-based Distributed Multimedia Transcoding Module (HadoopDMTM) and the streaming module 430 may be a Hadoop-based Distributed Multimedia Streaming Module Multimedia Streaming Module (HadoopDMSM).

As described above with reference to FIG. 3, the distributed multimedia streaming service providing system 300 may transcode content data. For example, the content data may be transcoded into an MPEG-4 video format so that it can be provided to various terminal devices 310 such as PCs, smart phones and tablet PCs.

The distributed multimedia streaming service providing system 300 integrates a Hadoop Distributed File System (HDFS) for storing content data and a MapReduce for parallel processing of distributed content data to provide a transcoding execution time .

The distributed multimedia streaming service providing system 300 can control a streaming operation using a streaming job distribution algorithm to reduce content delay and traffic bottlenecks.

In order to improve the overall performance of the distributed multimedia streaming service providing system 300, dual Hadoop clustering per physical cluster of the distributed multimedia streaming service providing system 300 may be used. Dual-Hadoop clustering can be used to distribute job tasks between transcoding and streaming.

The distributed multimedia streaming service providing system 300 can provide efficient distribution of content data and improved scalability through adherence to Hadoop policies.

The distributed multimedia streaming service providing system 300 can automatically perform a workflow of a sequential task for a streaming service deployment process.

The content data management server 410 may control and / or manage the transcoding module 420 and the streaming module 430. For example, the content data management server 410 may control and / or manage the transcoding operation of the content data performed by the transcoding module 420 and the streaming operation performed by the streaming module 430 of the content data. In addition, the content data management server 410 may monitor the resource use state of the distributed multimedia streaming service providing system 300. [

The content data management server 410 may include a web based dashboard module 402, a management server 405, and a data server 408.

The web-based dashboard module 402 includes a user interface (UI) for selecting options needed to monitor the usage rates of the dual Hadoop clusters in the streaming module 430 and the streaming servers of the streaming module 430. UI) to the terminal device 310. [

For example, the web-based dashboard module 402 may include a central processing unit (CPU) of a streaming server of the streaming module 430, a random access memory (RAM) and a streaming transmission rate to the terminal device 310 for the selection of an option associated with at least one of the options.

The web-based dashboard module 402 may also be used to select options for transcoding tasks that determine at least one of the resolution, bit rate, and frame rate of the content data. And can provide the UI to the terminal device 310.

The management server 405 can control and / or manage various jobs generated in each phase of transcoding and streaming of content data when the streaming service is provided to the terminal device 310. [ In addition, the management server 405 can monitor status information of jobs necessary for providing a streaming service such as uploading of content data, transcoding of content data, and transfer of content data to the terminal device. In addition, the management server 405 may use a streaming job distribution algorithm to distribute the load of the streaming servers of the streaming module 430.

The streaming task distribution may be performed using at least one of a round robin (RR) algorithm, a Least Connection (LC) algorithm, and a Streaming Resource based Connection (SRC) algorithm.

For example, the management server 405 may use the SRC algorithm for streaming work distribution. By using the SRC algorithm, at least one of the usage amount of the CPU of each streaming server 432, the usage amount of the RAM, and the streaming transfer rate can be considered in the distribution of the streaming work. Accordingly, the management server 405 can efficiently balance and distribute rapidly increasing streaming operations according to simultaneous connection of a plurality of users.

The management server 405 can manage operations related to transcoding of the content data of the transcoding module 420, transfer of content data, and extraction of thumbnail images from the content data, and the extracted thumb Manage jobs associated with tomorrow's images and / or information associated with the transcoded content data to the data server 408.

The data server 408 may include a content database 412 and a thumbnail database 414. The content DB 412 may store information related to content data registered from the streaming module 430 and the thumb tomorrow DB 414 may store thumb tomorrow images registered from the streaming module 430. [

The transcoding module 420 may transcode the content data into a predetermined format. The transcoding module 420 may apply options associated with transcoding to at least one of the display size, the codec method, and the container format that may be configured when transcoding the content data.

The transcoding module 420 may be more suitable for batch processing content data of large size collected over a predetermined period of time than processing small size content data collected in real time.

The transcoding module 420 may include one or more transcoders 422. The transcoders 422 may transcode the content data stored in the transcoding Hadoop cluster. The transcoders 422 may transcode the content data into a standard format for providing streaming service of the content data to the terminal device 310. [ For example, the standard format may be an MPEG-4 video format. The transcoders 422 may split and merge the content data, decode the encoded content data, change the size of the content data, and encode the decoded content data. Transcoders 422 may be MapReduce based transcoders.

The transcoding module 420 may also include an HDFS 424 for storing content data from a variety of sources. For example, the HDFS 424 may be used for distributed processing of content data to prevent excessive expenditure on communication such as transmission of content data during transmission of the content data. In addition, HDFS 424 may be used for large chunk size (64 MB) policies and / or user-level distributed systems suitable for processing content data. HDFS 424 may incorporate policies provided by the MapReduce for load balancing, fault tolerance, and distributed processing.

The transcoding module 420 may use the MapReduce for parallel processing of the content data and may use xuggler for transcoding the content data.

The transcoding module 420 may include a name node and one or more data nodes operating on the HDFS 424 of the transcoding Hadoop cluster. For example, a data node may act as a content storage server, and a name node may serve as a namespace server. Content data may be divided into one or more blocks and distributed to data nodes. The transcoding module 420 may manage the name node. In addition, the transcoding module 420 may manage and / or copy blocks of the partitioned content data and may transfer the content data to the streaming module 430. The content data transferred to the streaming module 430 may be transcoded content data.

The streaming module 430 may include a content storage server (not shown) that stores content data transcoded by the transcoding module 420. The streaming module 430 may transmit the transcoded content data stored in the content storage server to the terminal device 310 through streaming. For example, the streaming module 430 may operate on a streaming Hadoop cluster.

The streaming module 430 may further include one or more streaming servers 432, an HDFS 436, and a mountable HDFS 434. For example, the mountable HDFS 434 may be an HDFS 436. Streaming servers 432 may be distributed streaming servers based on MapReduce.

The streaming module 430 may include one or more data nodes and a name node operating on the HDFS 436. The transcoded content data may be divided into one or more blocks and distributed to the data nodes. The streaming module 430 may manage the name node. In addition, the streaming module 430 may manage and / or copy blocks of the segmented transcoded content data. For example, the streaming module 430 may recover content data and data nodes in the event of a failure of the data node and / or a loss of content data by copying the content data. That is, the streaming module 430 may use intelligent content data replication and recovery policies.

The most important factor in providing the streaming service to the user of the terminal device 310 may be that the users do not recognize the failure of the data node and / or the content loss and the provision of the seamless streaming service is maintained.

Unlike the distributed multimedia streaming service providing system 300, the existing streaming service providing system does not consider and / or introduce the duplication and automatic restoration policies of the content data. Therefore, the reliability of the streaming service providing system and the seamless streaming service Delivery may not be guaranteed. Also, in the existing streaming service providing system, when the transcoding operation and the streaming operation that require intensive computing resources are simultaneously performed, the quality of the streaming service provided to the user may not be guaranteed.

The technical contents described with reference to FIG. 1 and FIG. 3 may be applied as they are, so that a more detailed description will be omitted below.

FIG. 5 shows an SRC algorithm used by the management server shown in FIG. 4 for streaming work distribution.

5, the management server 405 receives at least one of a round robin (RR) algorithm, a Least Connection (LC) algorithm, and a Streaming Resource based Connection (SRC) Can be used to perform streaming job distribution.

The RR algorithm may stream the content data by selecting streaming servers in a predetermined order according to a video streaming request from the users. The LC algorithm may stream the content data by selecting a streaming server having the least number of streaming tasks.

The RR algorithm and the LC algorithm are intensive and require the CPU, memory, network bandwidth, and other resources required for components associated with massive physical resource usage, such as streaming processes, It is not suitable for processing streaming services in a distributed computing environment because it does not consider storage. Furthermore, since the quality of the streaming service is determined by variations in network conditions, task distribution methods that take into account only invariable elements are unsuitable. Thus, the RR and LC algorithms are only suitable for a static distributed computing environment and can handle simple text data and tasks based on web pages that do not require intensive resources. That is, the RR algorithm and the LC algorithm are limited because they put a heavy burden on the current Internet infrastructure and the streaming server.

The management server 405 may use the SRC algorithm for streaming work distribution. By using the SRC algorithm, at least one of usage of CPU, usage of RAM, and streaming rate of each streaming server can be considered in distribution of streaming work. For example, the amount of CPU usage and the amount of RAM usage indicate the CPU usage rate and the usage rate of RAM, and the streaming transfer rate may mean streaming transmission throughput.

Accordingly, the management server 405 can efficiently balance and distribute rapidly increasing streaming operations according to simultaneous connection of a plurality of users.

Linux command mpstat can be used to generate statistics on the CPU usage of servers. The free command is used to determine RAM usage, and proc / net / dev can be used to get the streaming transmission rate.

The parameters of the SRC algorithm may be as follows.

The ss _n _ID may be the ID of the streaming server 432 of the streaming module 430. The IDs of all streaming servers 430 may be denoted as ss _n _ID = ss ₁ , ss ₂ , ...., ss _n . ssu _n can be computed by adding the system usage rate for ss _n , CPU and RAM usage. The CPU utilization of ss _n is defined as scu _n and can be calculated by the Linux command mpstat. The RAM usage of ss _n is expressed as sru _n and can be calculated by free command. stt _n may be the current transmission throughput for one second generated by ss _n . The request ID may indicate the ID of the user who requested the streaming service.

To connect the optimal ss _n with the user streaming request, the SRC algorithm can calculate ssu _n and stt _n for each streaming server. ssu _n can be calculated by the following equation (1).

Further, stt _n can be calculated by the following equation (2).

current stt _n may be the total throughput of ss _n since the start of the streaming service. past stt _n is the total throughput of ss _n one second before the current measurement point.

The streaming task distribution sequence for selecting an optimal streaming server may be as follows. For example, an optimal streaming server may be a streaming server that is the highest idle state.

First, the management server 405 can distribute the current streaming task to the streaming server having the lowest value ssu _n .

Second, when there are two or more servers having the lowest value ssu _n , the management server 405 can distribute the streaming task to the streaming server having the lowest value stt _n .

The SRC algorithm selects a streaming server having optimal conditions for the streaming service among the streaming servers 432 by considering factors such as streaming network traffic and CPU and RAM usage rates .

Accordingly, the distributed multimedia streaming sub-providing system 300 can provide a rich and rich media streaming service through a streaming server having optimal conditions in a cloud computing environment having unpredictable network traffic. That is, the distributed multimedia streaming sub providing system 300 can improve the streaming task distribution capacity by providing the SRC algorithm.

In addition, the distributed multimedia streaming sub-providing system 300 can reduce content delay and prevent traffic bottlenecks while providing streaming services.

FIG. 6 illustrates a distributed multimedia streaming service providing system according to an example embodiment.

6 shows an example of the hardware configuration of the distributed multimedia streaming service providing system 300. [

6, the distributed multimedia streaming service providing system 300 includes a streaming task management module 510, a streaming server module (SSM) 520, and a HDFS-based A content storage server module (HSCSM) 530).

The STMM 510 is an embodiment of the management server 405 of the content data management server 410 of Figure 4 and the SSM 520 is an embodiment of the streaming server 432 of the streaming module 430 of Figure 4 And the HSCSM 530 may be one embodiment of the content storage server of the streaming module 430 of FIG.

The distributed multimedia streaming service providing system 300 may include 14 nodes running in a cloud computing cluster. The STMM 510 includes one node, the SSM 520 includes three nodes, and the HSCSM 530 may include ten nodes.

In FIG. 6, assume that the transcoding module 420 of FIG. 4 has completed the transcoding process for the content data, and skips the transcoding module 420.

The hardware specifications of each node can be summarized as shown in Table 1.

The SSM 520 may be streaming servers that run on NginX to stream the transcoded content to three nodes in the MPEG4 video format. HSCSM 530 may be the data nodes running on HDFS with ten nodes.

One node of STMM 510 may be a management server 405 for streaming test distribution using SRC.

The management server 405 of the STMM 510 can use the JSP to implement a web-based dashboard, use the swfuload 2.2.0.1 libraries for the content upload function, and use the Java library and / You can use the bash shell script and use the Google Chart API to display a graph of the monitoring system state.

The software specification for the distributed multimedia streaming service providing system 300 includes an H.264 streaming module 2.2.7 for streaming function, NginX 1.2.7 for streaming servers, and HDFS as a standard file system mounted on a Unix system mount, which allows you to mount fuse_dfs_0.1.0.

The STMM 510 is a major component of the distributed multimedia streaming service providing system 300, which plays an important role in efficient load balancing. Robust load balancing can be achieved by scheduling a streaming task distribution using the SRC algorithm.

First, a user can select a list of contents data by accessing a web dashboard of the STMM 510 using the terminal device 310. [

Second, when the user selects specific content data, the STMM 510 calculates the optimal streaming server (optimal streaming server) in the SSM 520 by calculating the usage rate of each streaming server and the current streaming transmission throughput streaming server.

Finally, the user can connect to the selected streaming server via the redirection mechanism for the HTTP request and receive the requested streaming service.

The SSM 520 may include streaming servers scheduled by the SRC algorithm of the STMM 510. As the number of streaming servers in the SSM 520 increases, the number of streaming tasks processed per streaming server decreases, so it may be possible to provide a rich multimedia streaming service to users.

However, the number of streaming servers may be less than the increase in physical configuration simply because of the total network throughput of the distributed multimedia streaming service providing system 300 and the physical (physical) configuration of the content storage servers of the HSCSM 530, May be determined in consideration of the physical configuration.

Each streaming server can use NginX with H.264 streaming module installed. NginX with H.264 modules is a typical web server and can be used to provide HTTP-based streaming services instead of streaming servers except for streaming protocols like RTP, RTSP, or RTMP. have.

In general, a particular streaming protocol can be selected based on the type of the target device when the streaming system is built. However, because streaming services must be allowed for most devices, HTTP-based streaming protocols are best suited can do.

HSCSM 530 may store transcoded content in data nodes in a distributed manner and may perform content replication and recovery when a data node failure or content loss occurs.

When providing a streaming service, the most important factor is the maintenance of a streaming service that is so necessary that the user can not notice the data node failure and content loss. Streaming service systems using the traditional approach do not include content replication management and automated recovery policy, resulting in system reliability and seamless service. Can not be. The HSCSM 530 can overcome these limitations by including Hadoop's core functionality and policies. In addition, mountable HDFS can be applied to HSCSM 530 so that HDFS can be mounted on the local file system. Accordingly, users can access a distributed file system having a complexity for use as a local file system by using general commands.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA) , A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

Claims (16)

Selecting an optimal streaming server among the streaming servers based on at least one of usage rate of physical resources and streaming transmission throughput; And
Streaming the content data using the optimal streaming server
/ RTI >
The method according to claim 1,
Wherein the selecting comprises:
Calculating a usage rate of physical resources of each of the streaming servers; And
Calculating a streaming transmission throughput of each of the streaming servers
/ RTI >
The method according to claim 1,
Wherein the selecting comprises:
Selecting a streaming server having the lowest usage rate of the physical resources among the streaming servers as the optimal streaming server
/ RTI >
3. The method of claim 2,
Wherein the selecting comprises:
Selecting a streaming server having the lowest streaming transmission throughput as the optimal streaming server among the streaming servers when the usage rate of the physical resource is the lowest value among the streaming servers
Further comprising the steps of:
3. The method of claim 2,
Wherein calculating the utilization rate of the physical resource comprises:
Calculating a usage rate of the physical resource based on the CPU usage rate and the RAM usage rate
/ RTI >
3. The method of claim 2,
Wherein the step of calculating the streaming transmission throughput comprises:
Calculating the streaming transmission throughput based on the current streaming transmission throughput and the previous streaming transmission throughput
/ RTI >
The method according to claim 6,
Wherein the current streaming transmission throughput is a total throughput since starting the streaming service.
The method according to claim 6,
Wherein the previous streaming transmission throughput is a total throughput up to one second before the point at which the current streaming transmission throughput is measured.
A streaming module including streaming servers; And
A content data management server for selecting an optimal streaming server among the streaming servers based on at least one of a usage rate of physical resources and a streaming transmission throughput
Lt; / RTI >
The streaming server may stream the content data
Distributed Multimedia Streaming Service Providing System.
10. The method of claim 9,
The content data management server comprises:
A usage rate of physical resources of each of the streaming servers, and calculates a streaming transmission throughput of each of the streaming servers.
10. The method of claim 9,
The content data management server comprises:
And selects a streaming server having the lowest usage rate of the physical resources among the streaming servers as the optimal streaming server.
12. The method of claim 11,
The content data management server comprises:
And selects a streaming server having the lowest streaming transmission throughput as the optimal streaming server among the streaming servers when there are two or more streaming servers having the lowest usage rate of physical resources.
11. The method of claim 10,
The content data management server comprises:
Wherein the usage rate of the physical resource is calculated based on a CPU usage rate and a RAM usage rate.
11. The method of claim 10,
The content data management server comprises:
And calculates the streaming transmission throughput based on a current streaming transmission throughput and a previous streaming transmission throughput.
15. The method of claim 14,
Wherein the current streaming transmission throughput is a total throughput since the start of the streaming service.
15. The method of claim 14,
Wherein the previous streaming transmission throughput is a total throughput up to one second before the point at which the current streaming transmission throughput is measured.

Images