KR101957239B1 - Method and apparatus for processing tasks - Google Patents

Abstract

Obtains host status information indicating a time required for each host to process a job and network status information indicating a status of a network with each host, Determining a host to process the job based on the network status information, and transmitting information necessary for processing the job to the determined host.

Description

[0001] The present invention relates to a method and apparatus for processing tasks,

The present invention relates to a method and apparatus for processing a job, and more particularly to a method and apparatus for processing a job in an SDN (Software-Defined Network) environment.

The conventional Internet architecture has been developed over the last 20 years, when client-server computing dominates, and adopts a hierarchical structure in which Ethernet switches and routers are arranged in a tree form. The explosive growth of mobile devices and content in recent years, the deployment of cloud-based virtualization services, and the study of future Internet deployments require a review of the current network structure and management structure. The main reason is due to the limitations of current network technology that can not keep pace with the rapidly changing network environment and the environment where vendors lead the technology development.

As a solution to the conventional Internet architecture, Software Defined Network (SDN) has recently been spotlighted. SDN is a next-generation networking technology that facilitates network routing, control and management of complex operations through software programming. For this purpose, the SDN separates the data plane and control plane of the network and provides a standardized interface between them. The network operator can program the control plane according to various situations, Can be controlled in various ways.

In SDN, it is very important to allocate tasks to hosts for efficient use of resources.

SUMMARY OF THE INVENTION An object of the present invention is to provide an apparatus and method for adaptively adjusting a period for checking a host status according to the degree of a task request.

In addition, it is an object of the present invention to provide a method and apparatus for processing a job efficiently based on a network situation and a host situation.

The problems to be solved by the present invention are not limited to the above-mentioned problem (s), and another problem (s) not mentioned can be clearly understood by those skilled in the art from the following description.

According to an aspect of the present invention, there is provided a method of processing a job, the method comprising: receiving information on a job to be processed from a client; Obtaining host status information indicating a time required for each host to process a job and network status information indicating a status of a network with each host; Determining a host to process the job based on the host status information and the network status information; And transmitting information necessary for processing the job to the determined host.

Wherein the obtaining of the information comprises: sending a monitoring message to each host to confirm the status of the network and each host; Receiving, in response to the transmission, a response message including information on the amount of waiting jobs and the average job processing time from each of the hosts; And obtaining the information based on the response message.

The step of transmitting the monitoring message may include adjusting a transmission period of the monitoring message based on the number of jobs waiting for the client.

The host status information may be determined based on information about an average task processing time, which is an average time required for completing a task in the host, and the number of tasks to be processed by the host.

The average job processing time may be determined by weighting the processing time of the recently processed job.

The weight may be determined based on the memory occupancy that the recently processed task occupies in the host.

The network status information may be determined based on average network delay information, which is an average time required for data transmission from the client to the host, a size of a job requested by the client, and a data loss rate.

Wherein the step of determining the host comprises the steps of: changing a weight of the host status information and the network status information based on characteristics of a job requested from the client; determining a host to process the requested job based on the changed weight; .

The step of determining the host may include: calculating a score for each host based on the host status information and the network status information; And stopping the processing of the job if the calculated score is not equal to or greater than the threshold.

The step of determining the host may further include, if there is a host having the calculated score equal to or greater than the threshold, requesting processing of the job to a host having no job waiting among the hosts having the calculated score exceeding the threshold .

The step of determining the host may further include, if there is a host having the calculated score exceeding the threshold, requesting processing of the task to a host having the highest score among the hosts having the calculated score equal to or higher than the threshold .

The step of transmitting information necessary for the task may include transmitting information necessary for processing the job to the router corresponding to the determined host.

The method includes: receiving a processing result of the job from the router; And transmitting the received processing result to the client.

The network may be SDN (Software-Defined Networking).

According to another aspect of the present invention, there is provided a task processing apparatus including: a receiving unit for receiving information on a task to be processed by a client; An information obtaining unit obtaining host status information indicating a time required for each host to process a job and network status information indicating a status of a network with each host; A determining unit that determines a host to process the job based on the host status information and the network status information; And a transmitting unit for transmitting information necessary for processing the job to the determined host.

The job processing apparatus according to an embodiment of the present invention quickly checks the status of the host when there are many waiting jobs by adjusting the transmission period of the monitoring message for checking the status of the host based on the number of waiting jobs Reduce data overhead when there are few waiting jobs.

In addition, the task processing apparatus according to an embodiment of the present invention improves the task processing speed by allocating tasks to the optimal host based on the host status information and the network status information.

1 shows a block diagram of a task processing system according to an embodiment of the present invention.
2 shows a block diagram of a work processing apparatus 200 according to an embodiment of the present invention.
3 is a diagram showing an example of an algorithm used to perform the monitoring step in the work processing apparatus 200 according to an embodiment of the present invention.
FIG. 4 is a diagram illustrating an example of an algorithm used to perform the job assignment step in the job processing apparatus 200 according to an embodiment of the present invention.
5 is a diagram illustrating an example of system status information (CSC) according to an embodiment of the present invention.
6 is a flowchart illustrating a job processing method according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various elements, but the elements should not be limited by the terms. For example, without departing from the scope of the present invention, a first component may be termed a second component, and similarly, the term " second component " The second component may also be referred to as the first component. The term < RTI ID = 0.0 > and / or < / RTI > includes any combination of a plurality of related listed items or any of the plurality of related listed items.

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.

The terminology used in this application is used only to describe a specific embodiment and is not intended to limit the invention. The singular expressions include plural expressions unless the context clearly dictates otherwise. In the present application, the terms "comprises" or "having" and the like are used to specify that there is a feature, a number, a step, an operation, an element, a component or a combination thereof described in the specification, But do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, 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 contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

Hereinafter, preferred embodiments according to the present invention will be described in detail with reference to the accompanying drawings.

1 shows a block diagram of a task processing system according to an embodiment of the present invention.

A work processing system 100 according to an embodiment of the present invention includes a client 101, a work processing apparatus 200, a router 103, and a host 104. [ The task processing system 100 according to an embodiment of the present invention may be a software-defined network (SDN) environment. In this case, unlike the conventional network, control functions such as job assignment and routing are performed by the job processing device 200 and the router 103 is controlled by the job processing device 200 and the host (104).

The client 101 requests the processing of one or more jobs through the job processing system 100. At this time, a plurality of clients 101 can be connected to the work processing system 100 at the same time.

The work processing apparatus 200 receives a job processing request from the client 101 and determines a host 104 to process each job. The work processing apparatus 200 manages a work list including information about jobs requested by the clients 101 and determines a host 104 suitable for processing the work based on the work list. The work processing apparatus 200 basically processes a first input first output (FIFO), that is, a job requested first, but exceptionally, a job that satisfies a specific requirement.

The router 103 receives the data necessary for the processing of the job from the work processing device 200 and transfers the data to the host 104 to be described later. When the work processing system 100 is implemented in the SDN environment, the router 103 performs only the function of transferring data to the host 104 under the control of the work processing apparatus 200. [ (Not shown) and a middleware (not shown) in a conventional network device in addition to the router 103. However, in the SDN environment, it is possible to control the operation of the work processing device 200 It will only perform the role of transmitting data.

The host 104 processes the data received from the router 103 and transmits the job processing result to the client 101 via the router 103. [

2 shows a block diagram of a work processing apparatus 200 according to an embodiment of the present invention.

The work processing apparatus 200 according to an embodiment of the present invention includes a receiving unit 210, an information obtaining unit 220, a host determining unit 230, and a transmitting unit 240.

The receiving unit 210 receives information on a job to be processed from the client 101. [ The receiving unit 210 can be requested to process a plurality of jobs from the plurality of clients 101. The job processing device 200 manages the waiting job list based on the information about the jobs received by the receiving unit 210 , And will determine the appropriate host to process the job based on the queued job list.

The information obtaining unit 220 obtains host status information and network status information.

The host status information is information indicating the time required for each host to process the task. The host status information can be determined based on the average job processing time at the host and the number of waiting jobs. For example, the host status information may be expressed as Expected Job Transfer Time (EJTT), and EJTT can be calculated by the following equation (1).

[Equation 1]

 EJTT = average job execution time × (number of waiting jobs -1)

In calculating the average task processing time, the most recently processed task may be weighted, where the weight is determined based on the degree of occupancy of the most recently processed task by the host's memory (or CPU) . Equation (2) is an example of the average work processing time.

&Quot; (2) "

W _new? W _old (1-a) + W _last? A

W _new : the current average working time of the host

W _old : the average time before the host was last run

W _lsat : Processing time of the last operation

a: memory occupancy when processing the last operation

The network status information can be determined based on the time required to transmit data from the client 101 to the host 104 and the packet loss rate. For example, the network state information may be expressed as an Expected Network Delay (ENT), and ENT may be calculated by Equation (3).

&Quot; (3) "

ENT = Average network delay * Number of packets in current job / (1-packet loss rate)

One task may be comprised of a plurality of packets, and the average network delay may be determined when a packet arrives at the host 104 from the client, such as processing, queuing, transmission, And a time required on the network up to < RTI ID = 0.0 > a < / RTI >

The host determination unit 230 determines a host to process the job based on the host status information and the network status information.

The host determining unit 230 may calculate the score for each host based on the host status information and the network status information. At this time, the host determining unit 230 may change the weight values of the host status information and the network status information according to the type of the job . For example, if a queued job is a network-sensitive job, it may give a higher weight to the network status information, and a higher weight may be given to the host status information if the queued job is sensitive to the processing speed of the data .

The host determining unit 230 compares the calculated host score with a predetermined criterion, and if there is no host satisfying the criterion, the host deciding unit 230 can stop processing the corresponding task until a host satisfying the criterion is found.

When there is a host satisfying the criterion, the host determining unit 230 can assign the task to the host having the highest score among the hosts satisfying the criterion, or assign the task to the host having no waiting task even if the score is low.

As an example, suppose that the score of the host calculated by the host determination unit 230 is as shown in Table 1 below.

Host ID Host Ranking (score) EJTT ENT Number of queued jobs A 1 (30) 20 10 2 B 2 (40) 20 20 0 C 3 (50) 30 20 One

In Table 1, for convenience of explanation, the host score is calculated according to the sum of EJTT and ENT, and it is assumed that the host score represents the ranking between the hosts. In this case, it is assumed that the smaller the sum of EJTT and ENT, the higher the ranking of the host. It is also assumed that the host determining unit 230 assigns a task only to a host whose sum of EJTT and ENT is 45 or less.

Accordingly, the host determination unit 230 assigns a task to either host A or host B. [ At this time, since the ranking of the host B is lower than that of the host A, but the host B does not have a waiting job, the host determining unit 230 can assign the task to the host B. In addition, according to the embodiment, the task can be assigned to the host A having a high ranking regardless of the number of waiting tasks.

The transmitting unit 240 transmits information necessary for processing the job to the host to process the job. The transmitting unit 240 may transmit information necessary for processing a job to a switch or a router corresponding to the host. In this case, information necessary for processing the job may be transmitted to the corresponding host via a switch or a router.

Hereinafter, the operation of each module of the work processing device 200 according to the time sequence will be described. For the sake of convenience of description, the monitoring step, the task allocation step, and the job processing step will be separately described.

[Monitoring step]

The transmission unit 240 periodically transmits a monitoring message to the host to confirm the status of the work processing system 100. [ At this time, the work processing apparatus 200 can control the transmission period of the monitoring message based on the number of waiting jobs. That is, the transmission period of the monitoring message is shortened (for example, 0.1 CP / sec) in order to quickly confirm the status of the job processing system 100 when the number of waiting jobs is equal to or higher than the threshold. On the other hand, when the number of queued jobs is below the threshold value, the transmission period is long (for example, 1 CP / sec) in a normal state, thereby reducing the system overhead.

An algorithm for the transmission of the monitoring message by the transmission unit 240 will be described later with reference to FIG.

The receiving unit 210 receives a response message from the host that has received the monitoring message. The response message may include information that can indicate the current state of the host. As an example, the response message may include the host ID, the average work time, and the number of wait jobs.

The information obtaining unit 220 obtains the network state information and the host state information based on the response message and updates the system (i.e., the cloud system) status information (hereinafter, referred to as CSC). The information obtaining unit 220 obtains the network status information based on the difference between the time at which the monitoring message was transmitted and the time at which the response message was received and the reception rate of the response message and stores the host status information on the basis of the information included in the response message Can be obtained. An example of the CSC updated by the information obtaining unit 220 will be described later with reference to FIG.

[Job Allocation Phase]

The host determination unit 230 determines a host to process the job based on the host status information and the network status information.

The host determining unit 230 calculates a score for each host based on the host status information and the network status information, and assigns the job to one of the hosts satisfying the criteria. An example of an algorithm regarding a process of determining a host to process a job in the host determination unit 230 will be described later with reference to FIG.

[Job processing step]

The transmitting unit 240 changes the destination IP to the IP of the host determined by the host determining unit 230 and transmits a connection request message. At this time, the source IP can still be maintained as the IP of the client.

The receiving unit 210 receives a connection acceptance message from the host.

The receiving unit 210 intercepts a message transmitted from the host to the client, changes the source IP to the IP of the work processing device 200, and forwards the source IP to the client. Accordingly, the client receiving the message will interpret the message as received from the work processing device 200 and send a response message (i.e., the formation of the session).

The transmitting unit 240 transmits to the host information (e.g., a packet necessary for the job) about the job requested by the client.

The host processes the corresponding job and transmits the processed job to the receiving unit 210.

The transmission unit 240 retransmits the job processing result to the client.

3 is a diagram showing an example of an algorithm used to perform the monitoring step in the work processing apparatus 200 according to an embodiment of the present invention.

2 line 310, it is checked whether the number of queued jobs (q) is equal to or larger than "0". If the number of waiting jobs is greater than zero, it means that there are jobs to be processed (or unallocated).

3 line 321 to line 322, it receives the response message and updates the system (i.e., the cloud system) status information (i.e., the CSC) based on the response message.

8 line 330, the transmission period of the monitoring message is set to 1.0 CP / sec.

Referring to the 9th line (342) to the 10th line (344), if the number of queued jobs is equal to or greater than the threshold value, the transmission period of the monitoring message is changed to 0.1 CP / sec. Accordingly, when the number of queued jobs is equal to or greater than the threshold value, the transmission period of the monitoring message is shortened to 1/10 so that the job processing device 200 can quickly recognize the status of the network and the host.

FIG. 4 is a diagram illustrating an example of an algorithm used to perform the job assignment step in the job processing apparatus 200 according to an embodiment of the present invention.

Referring to the first line 411, the job j to be processed is selected. The task to be processed can be selected according to the FIFO. That is, the requested operation can be selected first.

Referring to the second line 412, the score of the host to process the job is calculated. In FIG. 4, the score of the host is expressed by an average response time (ART) and is calculated based on the network status information and the host status information.

4 lines 422 to 424, it is determined whether or not there is a host having an ART below the threshold value. If there is no host having an ART below the threshold value, the task j is not allocated, RTI ID = 0.0 > ART. ≪ / RTI >

7 line 432 to line 434, if there is a host having no waiting job among the hosts having the ART below the threshold value, the task (j) is assigned to the host.

10 line 440, if there are jobs waiting for all hosts with an ART below the threshold, the task j is assigned to the most appropriate host (e. G., The lowest ART).

5 is a diagram illustrating an example of system status information (CSC) according to an embodiment of the present invention.

The CSC according to an embodiment of the present invention includes a time field 510, a C_ID field 520, a delay field 530, a loss field 540, an execution time field 550 and a waiting jobs field 560 .

The Time field 510 indicates the time when a message is received from the host.

The C_ID field 520 contains identification information of the host cloud.

The Delay field 530 indicates the expected delay time when transmitting data from the client to the host. The delay time can be determined based on the transmission / reception time of the monitoring message and the response message.

The Loss field 540 indicates the packet loss rate. The packet loss rate can be determined based on the transmission / reception ratio of the monitoring message and the response message.

The execution time field 550 indicates the time required to process the task at the host.

The Waiting Jobs field 560 indicates the number of jobs waiting on the host.

The Execution time field 550 and the Waiting Jobs field 560 play an important role in confirming the current state of the host.

Referring to FIG. 5, a host with ID "Host002" has the smallest delay time of "0.16" and the smallest packet loss rate of "0.04%". Also, the task execution time is the shortest with '2.11' and the number of waiting jobs is '5'. In particular, it can be confirmed that the status of the host is improved when compared with the response message received at 9:34:12:45. Accordingly, the host determination unit 230 can assign the task to the host "Host002 ".

6 is a flowchart illustrating a job processing method according to an embodiment of the present invention.

 In step s610, it is determined whether the number of jobs waiting in the job processing apparatus 100 exceeds a threshold value. If the number of waiting jobs exceeds the threshold value, the step s622 is performed, and if the number of waiting jobs is below the threshold value, the step s624 is performed.

In step s622, the monitoring period is shortened (for example, 0.1 CP / sec) and the monitoring message is transmitted to the host.

In step s624, the monitoring period is longer (for example, 1 CP / sec) and the monitoring message is transmitted to the host.

In step s630, a response message is received from the host.

In step s640, based on the response message, the network status information indicating the status of the network and the host status information indicating the status of the host are acquired, and the score for the host is calculated based on the obtained network status information.

In step s650, based on the host score, it is determined whether there is a host satisfying a specific criterion. If there is a host satisfying a specific criterion, step s660 is performed. Otherwise, step s650 is repeated to wait until there is a host satisfying the criterion.

In step s660, it is determined whether there is a host that does not have a waiting job among the hosts satisfying the specific criterion.

If there is a host that does not have a waiting job among the hosts satisfying the criteria, step s662 is performed to allocate the job to the host. However, if there are waiting jobs in all the hosts satisfying the criteria, step s664 is performed to allocate the jobs to the host having the highest score.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and / or features of the present invention, and how to accomplish them, will become apparent with reference to the embodiments described in detail below with reference to the accompanying drawings. It should be understood, however, that the invention is not limited to the disclosed embodiments, but is capable of many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, To fully disclose the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout the specification.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined by the scope of the appended claims and equivalents thereof.

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, but, on the contrary, Modification is possible. Accordingly, the spirit of the present invention should be understood only in accordance with the following claims, and all equivalents or equivalent variations thereof are included in the scope of the present invention.

101: Client
200: work processing device
210:
220: Information obtaining unit
230:
240:

Claims (15)

In the job processing apparatus, receiving information on a job to be processed from a client;
The method comprising: obtaining host status information indicating a time required for each host to process a job and network status information indicating a status of a network between the client and each host in the job processing apparatus;
In the work processing apparatus, determining a host to process the job based on the host status information and the network status information; And
And transmitting, in the job processing apparatus, information necessary for processing the job to the determined host,
The step of acquiring the information comprises:
Sending a monitoring message to each host to confirm the status of the network and each host;
Receiving, in response to the transmission, a response message including information on the amount of waiting jobs and the average job processing time from each of the hosts; And
And obtaining the host status information and the network status information based on the response message,
The step of transmitting the monitoring message may shorten the transmission period of the monitoring message if the number of waiting jobs is equal to or greater than the threshold and adjust the transmission period of the monitoring message to be longer when the number of waiting jobs is less than the threshold. To be processed. delete delete 2. The method of claim 1,
Based on information about an average task processing time which is an average time required for completing one task in the host and a number of tasks to be processed by the host. 5. The method according to claim 4,
Wherein a weight is assigned to a processing time of a recently processed work. 6. The method of claim 5,
Wherein the recently processed work is determined based on a memory occupancy rate in the host. 2. The method of claim 1,
The average network delay information which is an average time required for data transmission from the client to the host, the size of the job requested by the client, and the loss rate of the data. 2. The method of claim 1, wherein determining the host comprises:
Changing a weight of the host status information and the network status information based on a characteristic of the job requested from the client and determining a host to process the requested job based on the changed weight, To be processed. 2. The method of claim 1, wherein determining the host comprises:
Calculating a score for each host based on the host status information and the network status information; And
And suspending the processing of the job if the calculated score is not equal to or greater than a threshold value. 10. The method of claim 9, wherein determining the host comprises:
Further comprising the step of requesting a host having no job waiting among the hosts having the calculated score equal to or higher than a threshold value to process the job if the calculated score is equal to or greater than a threshold value. 10. The method of claim 9, wherein determining the host comprises:
And requesting processing of the job to a host having the highest score among the hosts having the calculated score equal to or higher than the threshold, if the host having the calculated score exceeds the threshold. The method as claimed in claim 1,
And transmitting information necessary for processing the job to the router corresponding to the determined host. 13. The method of claim 12,
Receiving a processing result of the job from the router; And
And transmitting the received processing result to the client. The network according to claim 1,
SDN (Software-Defined Networking). A receiving unit for receiving information on a job to be processed from a client;
An information obtaining unit for obtaining host status information indicating a time required for each host to process a job and network status information indicating a status of the network between the client and each of the hosts;
A determining unit that determines a host to process the job based on the host status information and the network status information; And
And a transfer unit for transferring information necessary for processing the job to the determined host,
The information obtaining unit obtains,
A monitoring message for confirming the status of the network and each of the hosts is transmitted to each of the hosts, and in response to the transmission, information about the amount of waiting jobs and the average job processing time from each host is included Wherein the host status information and the network status information are obtained based on the response message and the transmission period of the monitoring message is shortened when the number of waiting jobs is equal to or greater than the threshold, Wherein when the number of the monitoring messages is less than the threshold, the transmission period of the monitoring message is adjusted to be longer.

Images