KR20120036726A - Apparatus and method for controlling computing capacity for multiple computers sharing resources - Google Patents

Abstract

PURPOSE: A computing capacity control method and an apparatus thereof are provided to obtain computing resource by building cooperated cloud computing environment. CONSTITUTION: A request processing unit(101) receives a computer resource reserving or withdrawing request signal from a first computer. A decision making unit(110) decides suitability of the computing resource reserving or withdrawing request signal. If so, a computing resource virtualization manager(102) allocates the computing resource to the first computer. A computing capacity account manager(104) manages computing resource information to the first computer.

Description

TECHNICAL FIELD OF THE INVENTION A method and apparatus for adjusting computing capacity of a plurality of computers sharing a mutual resource {APPARATUS AND METHOD FOR CONTROLLING COMPUTING CAPACITY FOR MULTIPLE COMPUTERS SHARING RESOURCES}

The present invention relates to a method and apparatus for adjusting the computing capacity of a plurality of computers sharing mutual computing resources, and more particularly to a method and apparatus for adjusting the computing capacity that can build a cloud computing environment.

Recently, the use of cloud computing is expected to increase rapidly in many public and industrial fields. Cloud computing is a technology that integrates resources of computers in different physical locations into virtualization technology. The biggest advantage of cloud computing is that you can quickly adjust computing capacity without investing in additional assets. For example, a company that provides certain services to the outside world may fluctuate the demand for computing resources from time to time due to changes in customer preferences and seasonal factors. Can be.

However, the conventional cloud computing service providing environment (for example, Amazon EC2, Microsoft's Azure, etc.) has been a method in which one cloud computing service provider provides cloud computing services to multiple users.

This approach allows cloud computing service users to rely entirely on all computing services to cloud computing service providers, so that users who are accustomed to owning computing resources may experience loss of control over computing resources, security anxiety, and so on. It limits the spread of the environment.

In addition, many enterprises or organizations already own a large amount of computing resources for each purpose, and there is a need for a method and apparatus that can efficiently utilize such computing resources in a cloud computing environment.

On the other hand, unlike cloud computing services that rely entirely on computing service providers, there is grid computing in which multiple computers loosely share computing capacity. In a grid computing scheme, owners of each computing resource participating in grid computing may own their own computing resources while sharing the computing resources with each other as needed.

However, in the case of grid computing, a loose collaboration between computers (or between computer resource owners) does not guarantee how much computing resources are available to a computer resource owner at a given point in time, thus meeting business requirements. There is no limit to use for the service.

In particular, the computing capacity allocation of grid computing allocates and adjusts the computing capacity by considering only the work request and the available resource status predicted at the present time. However, the determination of computing resource allocation in a business environment should be able to predict future available computing capacity based on the computing resource owner's past contribution of computing resources.

Therefore, a computing resource control method and apparatus for collaboratively sharing computing resources while ensuring ownership of computing resources, such as grid computing, and ensuring computing resources available at a specific time, such as cloud computing services, are provided. need.

The present invention provides a method and apparatus for adjusting computing capacity that can adjust available computing capacity of each computer based on the computing resource demand and the computing resource sharing contribution of each computer among a plurality of computers sharing mutual computing resources.

Computing capacity adjustment method according to an aspect of the present invention comprises the steps of receiving a computing resource deposit request signal from a first computer; Determining suitability of the computing resource deposit request signal; Allocating the computing resource deposited by the first computer to a second computer if the computing resource deposit request signal is appropriate; And storing information about the computing resources deposited by the first computer and the computing resources allocated to the second computer in an account for the first computer and the second computer.

The method includes receiving a computing resource withdrawal request signal from the first computer; Determining suitability of the computing resource fetch request signal; And if the computing resource withdrawal request signal is appropriate, allocating computing resources of a third computer to the first computer.

The computing resource withdrawal request signal may include information about the size and time of the computing resource that the first computer requests permission of the other computer to use.

The method may further include storing information about computing resources allocated to the first computer in accounts for the first computer and the third computer.

The third computer may be a different computer from the second computer or may be the same computer.

The determining of the suitability of the computing resource drawing request signal may be determined by comparing the size of the computing resource previously deposited by the first computer with the size of the computing resource requiring permission of the other computer to use it.

The method may further include transmitting a signal requesting correction of the computing resource withdrawal request to the first computer if the computing resource withdrawal request signal is not appropriate.

The method may further include transmitting a signal for requesting correction of the computing resource deposit request to the first computer when the computing resource deposit request signal is not suitable.

The computing resource deposit request signal may include information about the size and time of computing resources that the first computer allows other computers to use.

The determining of the suitability of the computing resource deposit request signal may be determined by comparing a scale and a time of the computing resource that the first computer allows the other computer to use.

According to another aspect of the present invention, an apparatus for adjusting computing capacity may include: a request processing unit configured to receive a computing resource deposit request signal or a computing resource fetch request signal from a first computer; A decision unit determining a suitability of the computing resource deposit request signal or the computing resource fetch request signal; A computing resource virtualization manager configured to allocate the computing resource deposited by the first computer to another computer or to allocate the computing resource to the first computer when the computing resource deposit request signal or the computing resource withdrawal request signal is appropriate; And a computing capacity account manager that manages information about computing resources of the first computer and the other computer.

The apparatus may further include a computing capacity account database unit for storing information about computing resources in accounts for the first computer and the other computer.

The computing resource deposit request signal may include information about the size and time of computing resources that the first computer allows other computers to use, and the computing resource withdrawal request signal may be used by the first computer for other computers. It may include information about the size and time of computing resources requiring permission.

The decision unit may determine the suitability of the computing resource fetch request signal by comparing the scale of the computing resource previously deposited by the first computer with the scale of the computing resource requiring the other computer to permit use.

The decision determiner may determine suitability of the computing resource deposit request signal by comparing the size and time of a computing resource that the first computer allows other computers to use.

When the decision determiner determines that the computing resource withdrawal request signal is not suitable, the request processing unit may transmit a signal for requesting correction of the computing resource withdrawal request to the first computer.

When the decision determiner determines that the computing resource deposit request signal is not suitable, the request processor may transmit a signal requesting correction of the computing resource deposit request to the first computer.

Computing capacity control devices and methods can be used to create collaborative cloud computing environments, and as a result, each computer can cope with fluctuations in computing capacity while retaining ownership of computing resources. In addition, the idle computing capacity of each computer that is unused and volatilized can be utilized, thereby increasing the efficiency of the overall resource usage.

1 illustrates an example of computing resource usage of each computer having its own computing resource.
2 conceptually illustrates a computing capacity adjustment method in accordance with the present invention.
3 illustrates a configuration of a computing capacity adjusting apparatus according to an embodiment of the present invention.
4 illustrates a method of processing a computing capacity deposit request in a computing capacity adjusting device.
5 illustrates a method of processing a computing capacity withdrawal request in a computing capacity adjusting apparatus.
6 illustrates an application of the present invention in a collaborative cloud computing environment.

1 illustrates an example of computing resource usage of each computer having its own computing resource.

Referring to FIG. 1, computers A, B, and C are computers that independently possess computing resources, and computers A, B, and C have the same available computing capacity (measured as CPU processing capacity, memory capacity, etc.). There may be. However, the time series pattern of computing capacity requirements for a particular time period may be different. For example, as shown in FIG. 1, computer A may exceed the available computing capacity held by the computing capacity requirements concentrated in the second half of the specific time interval. Computer B may exceed the available computing capacity by concentrating computing capacity requirements in the first half of the particular time interval. Computer C may exhibit some degree of computing capacity requirements.

In this case, computer C can perform smooth computing tasks, while computers A and B experience severe delays in computing tasks at specific times, that is, at certain times when computing capacity requirements are greater than their computing capacity. Done.

In other words, each computer may have a computing capacity demand exceeding a computing capacity in some cases, and in this case, there may be a delay in computing work. To solve this, each computer needs to increase its own computing capacity, but this is inefficient in terms of cost and wasteful in terms of resource usage.

However, assuming that the computing capacity and computing capacity requirements of computers A, B, and C are integrated and operated, the combined integrated computing capacity requirements of computers A, B, and C are compared to the computing capacity requirements of one computer. The probability of smoothing is increased and the probability of being a small amount compared to unified computing capacity increases. This increases the probability that computers A, B, and C will be able to ensure smooth computing tasks at all times.

In FIG. 1, only three computers are illustrated for convenience of description, but when the computing capacity requirements of more computers are added together, the integrated computing capacity requirements are smoothed in time by a statistical 'the law of large number'. The chances are very high.

2 conceptually illustrates a computing capacity adjustment method in accordance with the present invention.

Referring to FIG. 2, computer A may take control of some computing capacity so that the surplus computing capacity (denoted as control transfer computing capacity) is available to other computers during a period when less computing capacity is required than available computing capacity. It delivers to and takes control of the computing capacity of any other computer (represented as control receiving computing capacity) in a period in which a computing capacity exceeding the held computing capacity is required.

Similarly, computer B receives and takes control of the computing capacity of another computer while the computing capacity requirements are concentrated. In addition, computer B transfers the computing capacity control right in the period in which the computing capacity requirement is reduced by the used computing capacity.

In an environment where multiple computers are connected, control is given to other computers by explicitly declaring how long and how long they can use some of their surplus computing capacity, excluding the computing capacity required by computing capacity. The handing over is referred to hereinafter as depositing (or paying, the same below) computing capacity for convenience.

After explicitly declaring the amount of computing capacity and the amount of computing capacity that a particular computer will use, the transfer of control of computing capacity from another computer is referred to as the withdrawal (or loan) of computing capacity for convenience.

In addition, the computing capacity summed up by computing the amount of computing capacity deposited by a specific computer as a negative value is expressed as the balance of the computing capacity of the computer.

The reason for explicitly declaring the usage and size of computing capacity to deposit / draw and manage the balance of computing capacity is to allow each computer to predict the available computing capacity at any given time, while at the same time This is to allow the current computing resource usage to be determined according to the capacity sharing contribution.

In order to implement the above-described process of depositing / withdrawing the computing capacity and managing the balance, there is a need for a computing capacity adjusting apparatus and method capable of processing the deposit / withdrawal request of computers with computing resources and virtually distributing computing resources to each computer. Do.

3 illustrates a configuration of a computing capacity adjusting apparatus according to an embodiment of the present invention.

Referring to FIG. 3, the computing capacity adjusting device 100 includes a request processing unit 101, a computing resource virtualization managing unit 102, a decision making unit 110, a computing capacity account managing unit 104, and a computing capacity accounting database 103. , data base, DB).

The request processing unit 101 receives a computing capacity deposit or withdrawal request signal from an individual computer (or automated software installed in the individual computers) and transmits the result to the computing resource virtualization management unit 102 and the decision unit 110. Module. Herein, the individual computer may be referred to as a computing resource holder.

The decision unit 110 receives a computing capacity deposit or withdrawal request signal from the request processing unit 101 and determines a possibility of depositing or withdrawing the computing capacity for a computer that has performed the request.

The decision maker 110 may include a virtualization scheduler 111 and a deposit withdrawability determination unit 112. The virtualization scheduler 111 determines an optimal virtualization scheme by looking for virtual organizational alternatives for each computing resource that satisfy an already in progress or newly requested computing capacity deposit or withdrawal request of each computer.

The deposit withdrawability determination unit 112 compares the computing capacity deposit or withdrawal request with the computing capacity account balance of the computing resource holder to determine whether to approve the deposit and withdrawal of the computing capacity. That is, the computing capacity available to each computer at a certain point of time may be determined by the deposit withdrawability determination unit 112 according to the past contribution of the computing capacity sharing of each computer.

The computing resource virtualization management unit 102 is a module that allocates and adjusts computing resources to a server equipped with a virtualization OS according to a deposit or withdrawal decision of the decision maker 110. Here, the server may be recognized as a device in the individual computer or connected to the individual computer by the virtualization OS, and the user may perform a desired computing task through the server.

The computing capacity account manager 104 is a module for managing an account balance according to computing capacity deposits, withdrawals, etc. of the computing resource holders. That is, each computing resource holder may have a unique account, and the computing capacity account manager 104 may manage the computing capacity deposited (paid) or withdrawn (loaned) by the computing resource holders for each account.

The computing capacity account DB 103 is a memory for storing the computing capacity account. The computing capacity account DB may store information, such as an ID unique to each computer, a computing capacity balance according to a deposit or withdrawal result, and the like.

4 illustrates a method of processing a computing capacity deposit request in a computing capacity adjusting device.

Referring to FIG. 4, the request processor 101 receives a deposit request signal of computing capacity (S110). The deposit request signal may include information about the ID of the computer, the amount of computing capacity to deposit, and the deposit time. The request processor 101 transmits the deposit request signal of the computing capacity to the decision unit.

The decision unit 110 determines an appropriateness of the computing capacity deposit request (S120). For example, the decision unit 110 may compare the predetermined threshold value with the information of the computing capacity to be deposited to determine the appropriateness. For example, the deposit request may be rejected when the deposit time for allowing the use of the control right over the memory of the computer A in the computing capacity requested by the deposit is shorter than the threshold or the amount of memory to be deposited is smaller than the threshold. In such a case, frequent allowance of virtualization schemes would be required if the deposit request was allowed. Accordingly, the deposit withdrawability determination unit 112 of the decision unit 110 may determine that the deposit request is suitable only when the deposit computing capacity and the deposit period are larger than the predetermined threshold. The allowable threshold value of the deposit time or the computing capacity scale may be set to a predetermined value or may be defined as a value that varies depending on the situation.

When the deposit withdraw possibility determination unit 112 determines that the deposit request is not suitable, the request processing unit 101 transmits the computing capacity deposit request correction signal to the computer that requested the deposit (S130).

When the deposit withdrawability determination unit 112 determines that the deposit request is a suitable deposit request, the deposit request signal is transmitted to the virtualization scheduler 111, and the virtualization scheduler 111 may use the computing capacity of the deposit request computer to another computer. Coordinate future virtualization programming schedules. The computing resource virtualization management unit 102 changes and executes the virtualization program by allocating computing resources to each server according to the virtualization program schedule through the virtual OS of the server (S140).

The computing capacity account management unit 104 changes the computing capacity account balance according to the changed virtualization program (S150). For example, if computer A deposits computing capacity M, and computer B borrows the computing capacity M, the computer capacity account balance is stored by updating and storing this information in each computing capacity account for computers A and B. Change it.

That is, in the computing capacity account of a specific computer, information about the computing capacity that the specific computer has contributed (deposited or paid) in the past, the computing capacity used in the past (withdrawal or loan) is stored and managed.

The total amount of deposit for computing resources can increase rapidly over time. To prevent this, you can set the computing capacity to be deposited by a particular computer so that it does not exceed a certain value even if time passes indefinitely. For example, if computer A continuously deposits computing capacity by C, a rate of α <1 may be applied, but the total amount of computed deposited may not exceed C / (1-α) even if time passes indefinitely. . This means that the value of future available computing capacity is much greater than the available computing capacity of the past.

5 illustrates a method of processing a computing capacity withdrawal request in a computing capacity adjusting apparatus.

Referring to FIG. 5, the request processing unit 101 receives a computing capacity drawing request signal (S210). The withdrawal request signal may include ID, computing capacity information, and withdrawal time information of each individual computer. Then, the request processor 101 transmits the withdrawal request signal to the decision unit 110.

The deposit withdrawability determination unit 112 of the decision unit 110 requests and receives the computing capacity account balance of the withdrawal request computer from the computing capacity account management unit 104, and checks whether the computing capacity account balance is sufficient to process the withdrawal request. It is determined whether or not (S220).

For example, when the computing capacity account balance is insufficient, that is, when the withdrawal request is too large compared to the computing capacity account balance, the withdrawal request is inappropriate (S250). That is, the request processing unit 101 transmits a correction signal of the computing capacity drawing scale or the drawing time to the computer which requested the drawing.

For a withdrawal request with sufficient computing capacity account balance, a withdrawal reservation for the computing capacity balance of the computer account is first performed in order to prevent double withdrawal of computing resources (S230). Information about the withdrawal scheduled computing resources is transmitted to the virtualization scheduler 111 to adjust the virtualization schedule (S240). The computing resource virtualization management unit 102 changes and executes a virtualization program of each server by allocating computing resources according to a virtualization program schedule through a virtual OS of a related server. That is, unlike the related art, when a specific computer requests computing capacity from another computer in an environment where a plurality of computers are connected, the appropriateness of the request may be determined based on the past computing capacity contribution of the specific computer. There is a difference.

The computing capacity account management unit 104 changes the computing capacity account balance according to the changed virtualization program (S270).

The computing capacity adjusting method described above with reference to FIGS. 4 and 5 has been described using the deposit / withdraw of computing capacity as an example. The loan processing of the computing capacity may be performed in the same manner as the computing capacity withdrawal processing. For example, the suitability of the loan request for computing capacity may be determined based on a past computing capacity contribution of the requested computer, or may be determined based on a predetermined computing capacity threshold. In addition, the computing capacity payment processing may be performed similarly to the computing capacity deposit processing. The appropriateness of the computing capacity payment request may be determined based on a predetermined computing capacity threshold.

6 illustrates an application of the present invention in a collaborative cloud computing environment.

Referring to FIG. 6, one logical server farm is used in a distributed space where computing resources owned by multiple organizations are physically in the same space, such as an Internet data center (IDC), or in a distributed space where communication complexity between computing resources is not a big problem. ) Each organization's computing resources can consist of one computing server or a group of computing servers.

Computing resources of each organization may be connected to the computing capacity adjusting device to request deposit / pay for idle computing capacity, or to withdraw / loan the computing capacity requirement, thereby expanding or reducing virtually available computing capacity.

According to the present invention, an idle computing capacity of one organization may be explicitly deposited for a certain period of time so that another organization may use it, and then the available computing capacity may be withdrawn at a specific time point. Alternatively, the excess computing capacity required by one organization may be explicitly borrowed and used by another organization's computing capacity, and then the computing capacity loaned for a specific period may be made available to users of another organization.

Through the computing capacity adjusting method according to the present invention, each computer can not only cope with the fluctuation of computing capacity intentionally while possessing the ownership of computing resources, but also use the idle computing capacity of each organization which is volatilized because it is not used. Available, increasing the efficiency of resource utilization across the board.

In other words, organizations in a trusted relationship can work together to create a collaborative cloud computing environment that enables rapid expansion of computing capacity when needed, such as a cloud computing environment, without losing control of their computing resources.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made therein without departing from the spirit and scope of the invention. You will understand. Therefore, the present invention is not limited to the above-described embodiment, and the present invention will include all embodiments within the scope of the following claims.

Claims (19)

In the computing capacity adjustment method of a plurality of computers sharing mutual resources,
Receiving a computing resource deposit request signal from a first computer;
Determining suitability of the computing resource deposit request signal;
Allocating the computing resource deposited by the first computer to a second computer if the computing resource deposit request signal is appropriate; And
And storing information about the computing resources deposited by the first computer and the computing resources allocated to the second computer in an account for the first computer and the second computer. How to adjust the computing capacity of a plurality of computers. The method of claim 1,
Receiving a computing resource withdrawal request signal from the first computer;
Determining suitability of the computing resource fetch request signal; And
And allocating computing resources of a third computer to the first computer if the computing resource withdrawal request signal is appropriate. 3. The plurality of computers sharing mutual resources according to claim 2, wherein the computing resource withdrawal request signal includes information on the size and time of computing resources for which the first computer requires permission from another computer. How to scale computing capacity. 3. The plurality of shared resources of claim 2, further comprising storing information about computing resources assigned to the first computer in accounts for the first computer and the third computer. How to adjust the computing capacity of your computer. The method of claim 4, wherein the third computer is a different computer from the second computer. 5. The method of claim 4, wherein the third computer is the same computer as the second computer. The method of claim 4, wherein the determining of the suitability of the computing resource withdrawal request signal comprises: comparing the size of the computing resource previously deposited by the first computer with the size of the computing resource requiring the other computer to allow use. Characterized in that the computing capacity control method of a plurality of computers sharing mutual resources. 3. The method of claim 2, further comprising: transmitting a signal for requesting correction of a computing resource withdrawal request to the first computer if the computing resource withdrawing request signal is not appropriate. To adjust the computing capacity of your computer. The method of claim 1, further comprising: if the computing resource deposit request signal is not suitable, transmitting a signal for requesting correction of the computing resource deposit request to the first computer. To adjust the computing capacity of your computer. The method of claim 1, wherein the computing resource deposit request signal includes information on the size and time of computing resources that the first computer allows other computers to use. How to adjust computing capacity. The method of claim 1, wherein the determining of the suitability of the computing resource deposit request signal comprises determining the size and time of the computing resource that the first computer allows the other computer to use in comparison with a predetermined threshold. Computing capacity adjusting method of a plurality of computers sharing mutual resources. A computing capacity adjusting device,
A request processor to receive a computing resource deposit request signal or a computing resource withdraw request signal from the first computer;
A decision unit determining a suitability of the computing resource deposit request signal or the computing resource fetch request signal;
A computing resource virtualization manager configured to allocate the computing resource deposited by the first computer to another computer or to allocate the computing resource to the first computer when the computing resource deposit request signal or the computing resource withdrawal request signal is appropriate; And
And a computing capacity account manager configured to manage information about computing resources of the first computer and the other computer. 13. The apparatus of claim 12, further comprising a computing capacity account database unit for storing information about computing resources in accounts for the first computer and the other computer. 13. The apparatus of claim 12, wherein the computing resource deposit request signal includes information about the size and time of computing resources that the first computer allows other computers to use. 13. The apparatus of claim 12, wherein the computing resource withdrawal request signal includes information about the size and time of computing resources that the first computer requests to allow other computers to use. The method of claim 12, wherein the decision unit is further configured to determine the suitability of the computing resource fetch request signal by comparing the computing resource size previously deposited by the first computer with the size of the computing resource requiring the other computer to allow use. A computing capacity adjusting device. The method of claim 12, wherein the decision unit determines the suitability of the computing resource deposit request signal by comparing the size and time of the computing resources that the first computer allows the other computer to use the predetermined threshold value. Computing capacity control device. The computing capacity of claim 12, wherein when the decision determiner determines that the computing resource withdrawal request signal is not suitable, the request processing unit transmits a signal for requesting correction of the computing resource withdrawal request to the first computer. Regulating device. The computing capacity of claim 12, wherein when the decision determiner determines that the computing resource deposit request signal is not appropriate, the request processor transmits a signal to the first computer to request a correction of the computing resource deposit request. Regulating device.

Images