WO2010032501A1 - 情報処理システムの運用管理方法 - Google Patents
情報処理システムの運用管理方法 Download PDFInfo
- Publication number
- WO2010032501A1 WO2010032501A1 PCT/JP2009/054446 JP2009054446W WO2010032501A1 WO 2010032501 A1 WO2010032501 A1 WO 2010032501A1 JP 2009054446 W JP2009054446 W JP 2009054446W WO 2010032501 A1 WO2010032501 A1 WO 2010032501A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- information processing
- power
- air conditioner
- workload
- power consumption
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/16—Constructional details or arrangements
- G06F1/20—Cooling means
- G06F1/206—Cooling means comprising thermal management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
- G06F1/26—Power supply means, e.g. regulation thereof
- G06F1/32—Means for saving power
- G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
- G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/46—Multiprogramming arrangements
- G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
- G06F9/5094—Allocation of resources, e.g. of the central processing unit [CPU] where the allocation takes into account power or heat criteria
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Definitions
- the present invention relates to an operation management method for an information processing system including a group of information processing apparatuses such as a server apparatus, a storage apparatus, and a network apparatus, and a cooling facility for cooling the information processing apparatus.
- the present invention relates to assignment of a workload to an information processing apparatus group necessary for realizing operation.
- information processing devices such as server devices, storage devices, and network devices are expected to increase explosively toward the rapid development of information communication in recent years and the era of the fusion of broadcasting and communication.
- the cooling power that is consumed by the cooling equipment that cools it also accounts for a large proportion. Therefore, the reduction of the total power that combines the device power and the cooling power has become an issue.
- Measures for information processing devices such as server devices, storage devices, and network devices, cooling facilities, and operation management have begun to save power in data centers.
- improvement in performance per power consumption by low power devices and adoption of a power saving function by switching operation states according to a work load are being promoted.
- cooling facilities air conditioners themselves have improved operating efficiency, optimized airflow design at the air supply and exhaust ports, and introduced local and liquid cooling.
- operation management operational information monitoring, job scheduling, and improvement of operation efficiency and consolidation of devices by virtualization are listed as major measures.
- the cooling facility is generally designed based on the maximum rated power of the device group, the power consumption of the cooling facility during operation greatly depends on the distribution and fluctuation of the power of the device group with respect to the cooling facility.
- the efficiency of the air conditioner depends on the power of the apparatus, that is, heat generation, the positional relationship and distance between the air conditioner and the apparatus, the temperature, the air volume, the wind direction, and the like, so the cooling power is greatly affected by the power and arrangement of the apparatus.
- server device consolidation will be used to save power using virtualization technology due to advances in data center operation management technology.
- measures to reduce the total power of the equipment group and the cooling equipment are necessary, taking into account the time variation.
- operation management methods for servers and storage devices such as server devices and storage devices.
- local or sequential measures are taken. Stays.
- Patent Document 1 in a management server that performs job scheduling for a parallel computer, a new job is input to a computer having a low temperature based on the temperature sensor information of the computer. The job is moved from the computer to the low-temperature computer to prevent the parallel computer from being damaged by high temperatures and performance degradation. In addition, before and after the movement of the job, whether or not the movement is possible is determined by estimating the power consumption of each computer and the cooling device provided for each computer based on the temperature information.
- Patent Document 2 in a management system for a plurality of computers, a superheated computer and a non-superheated computer are extracted based on computer temperature distribution and operation information, and the software is moved from the former to the latter. To save power. Further, after extracting the target computer, the computer power before and after the movement of the software and the fluctuation of the air conditioning power are compared to determine whether or not the movement is possible. The computer power is obtained from the operation information, the air conditioning power is obtained from the temperature distribution, and the temperature distribution is obtained from the temperature sensor, the temperature history, and the operation information.
- Patent Document 3 in a method of assigning workload to a plurality of servers in a data center, a requested workload profile is compared with a history profile, and a server The requested workload is assigned to the server according to the history of the lowest power for the air conditioning, and randomly assigned if there is no matching history.
- the history profile includes the server location, class, operation information, incoming air temperature, exhaust temperature, workload type, and server and air conditioning power.
- the power of the server and air conditioning is obtained from the inlet / outlet temperature of the server, the specific heat, and the air volume, or is measured by a power meter.
- Patent Document 4 in a method of distributing power to a plurality of servers in a data center, an ideal analog temperature distribution, that is, power distribution is approached. Lending and borrowing power budgets between servers or racks in close proximity to each other and specifying the discretized power state of servers based on the budget allocation prevents server failures due to hot spots and cold spots.
- the thermal multiplier indicating the ideal power for each server is obtained from the exhaust temperature of each server, the reference exhaust temperature of the average server, and the supply temperature of the air conditioning. JP 2004-126968 A JP 2007-179437 A US Patent Application Publication No. 2006/0259621 US Patent Application Publication No. 2006/0259793
- Patent Document 1 a job is assigned to a computer having a low temperature among the parallel computers, but this does not necessarily save power.
- the power consumption of the computer is estimated before and after the job is moved, the power related to the movement source and the movement destination is reduced, but this is only a local power saving for the entire parallel computer. That is, it does not lead to a total power saving in the computer room.
- the computer power and air conditioning power fluctuations are compared before and after the movement of the software, but the comparison target is limited to the overheated and non-heated computers extracted, and the computer group and air conditioning equipment are combined. The total power is not considered. That is, it does not lead to a total power saving in the computer room.
- Patent Document 3 since the workload is assigned to the server based on the history profile in which the power of the server and the air conditioning is the lowest, the power combined with the server group and the air conditioning equipment is reduced for the workload within the history range. Although there is no history, it can be improved only sequentially according to the history accumulation. That is, it is necessary to take a long time to realize power saving in the computer room. In Patent Document 4, since a power budget is lent and borrowed between adjacent servers so as to approach an ideal temperature distribution, the total power of the entire server group is not reduced. That is, it does not lead to a total power saving in the computer room.
- the conventional technology has a problem that only local power saving can be realized, or it takes a long time to realize power saving.
- An object of the present invention is to realize power saving of a computer room in a short time by optimizing work load allocation for the information processing apparatus in a computer room composed of an information processing apparatus and an air conditioner.
- the present invention has the following features in order to realize power saving of a computer room by optimizing work load allocation in a computer room composed of an information processing device and an air conditioner.
- a feature of the representative embodiment is that, for an air conditioner, an air conditioner operating coefficient related to the information processing apparatus is calculated.
- the air conditioner operating coefficient COP is the heat exchange efficiency of the air conditioner when the air conditioner is operated in order to exchange heat in the computer room, and is represented by (Equation 1).
- COP Heat exchange capacity / Air conditioner power (1)
- the air conditioner operating coefficient related to the information processing device of the air conditioner is a COP when the air conditioner exchanges heat generated by the specific information processing device, and is referred to as a device-specific COP here.
- COPji which is a device-specific COP related to the information processing device i of the air conditioner j, can be expressed by (Expression 2) using the calorific value PDi of the information processing device i and the air conditioner power PCji related to the information processing device i of the air conditioner j.
- a certain COPji can be expressed by (Expression 3) using a distance Lji between the air conditioner j and the information processing apparatus i, an air conditioner operation coefficient COPj unique to the air conditioner j, and an environmental constant A.
- the air conditioner operating coefficient COPj can be obtained from the specification of the air conditioner j, and the distance Lji can be obtained from arrangement information including the positions of the air conditioner and the information processing device group in the computer room.
- the air conditioner power related to a certain information processing device can be calculated by the device-specific COP.
- the heat generation amount of the information processing apparatus is assumed to be equal to the apparatus power, and the air conditioner power PCji related to the information processing apparatus i of the air conditioner j is Using the device power PDi and COPji of the device i, it can be expressed by (Equation 4).
- the air conditioner power PCi related to the information processing apparatus i is such that the amount of heat generated by the information processing apparatus processed by the air conditioner is proportional to the COPji ratio of each air conditioner.
- N, PDi, COPji can be used to express (Equation 5).
- Yet another feature is to create a device-related power consumption formula that is the sum of the device power and the air conditioner power for each information processing device.
- the apparatus-related power consumption type Pi (x) of the information processing apparatus i can be expressed by (Equation 6) using the apparatus power type PDi (x) of the information processing apparatus i and the air conditioner power PCi (x) related to the information processing apparatus i. .
- Pi (x) PDi (x) + PCi (x) (Expression 6)
- PDi (x) and PCi (x) are functions of the resource utilization rate x of the information processing apparatus.
- the resource usage rate is mainly a CPU usage rate or a memory usage rate.
- the resource utilization rate is mainly the disk area utilization rate.
- Pi (x) is the total of the apparatus power and the air conditioner power, and the air conditioner power takes into account the positional information of the air conditioner and the information processing apparatus group in the computer room. That is, by using Pi (x), it is possible to evaluate the power consumption in consideration of the position information in the computer room without being aware of the position information in the computer room.
- the power consumption of the entire computer room is calculated from the device-related power consumption formula of each information processing device. Since the device-related power consumption formula Pi (x) of the information processing device i includes the device power and the air conditioner power related to the corresponding device, the total of the device-related power consumption formula Pi (x) of each information processing device in the computer room is the sum. By doing so, it is possible to obtain the power consumption of the entire computer room.
- the power consumption P (x) of the entire computer room can be expressed by (Equation 7).
- the workload allocation is determined so as to realize the power saving of the entire computer room using the power saving evaluation index based on the device related power consumption formula of the information processing apparatus.
- the workload in the server device includes an operation of an application or a virtual OS or a response to a connected client.
- the work load of the storage device is a disk storage area or the like.
- the work load of the network device is network throughput or the like.
- the power saving performance evaluation index is the processing performance per power consumption when the resource usage rate of the information processing apparatus is the maximum value of the resource usage rate.
- the processing performance in the server device is, for example, a benchmark performance value such as SPEC or the number of transactions per unit time.
- the processing performance in the storage device is, for example, a disk storage area.
- the resource utilization rate maximum value is the maximum value of the resource utilization rate of the information processing apparatus, and is determined by the administrator based on the utilization policy of the information processing apparatus.
- the power saving performance evaluation index EEi regarding the information processing apparatus i assumes that the processing performance of the information processing apparatus i is proportional to the resource utilization rate, and uses the resource utilization maximum value Xi and the processing performance maximum value Wi of the information processing apparatus i. (Expression 8)
- EEi Xi ⁇ Wi / Pi (Xi) (Equation 8)
- EEi is a constant, and the larger the value, the better the performance of the information processing apparatus i versus the power consumption performance.
- the power consumption shown here is the total value of the apparatus power and the air conditioner power related to the information processing apparatus i. That is, by assigning a work load in preference to the information processing apparatus i having a large power saving evaluation index EEi, it is possible to assign a work load that minimizes the power consumption of the entire computer room.
- the resource usage rate of an information processing device with a large power-saving evaluation index approaches the maximum resource usage rate, and the resource usage rate of an information processing device with a small power-saving evaluation index becomes zero.
- the approaching bipolar phenomenon appears.
- the power saving performance evaluation index EEi evaluates the power saving performance when the resource usage rate of the information processing apparatus is the maximum value of the resource usage rate in consideration of this bipolar phenomenon.
- Yet another feature is to control the workload that operates in each information processing device based on the workload allocation result.
- the information processing apparatus group has a workload control agent, and the workload control agent controls the workload to be operated in the information processing apparatus according to the workload assignment result.
- the air conditioner power is calculated from the device-specific COP and the device power measurement value of the information processing device.
- the air conditioner power PCj of the air conditioner j is assumed that the heat generation amount of the information processing device processed by the air conditioner is proportional to the COPji ratio of each air conditioner, and the number N of air conditioners, the device power measurement value PDi of the information processing device i, It can be expressed by (Equation 9) using COPji.
- each air conditioner is controlled based on the calculation result of the air conditioner power.
- the air conditioner has an output control agent, and the output control agent controls the output of the air conditioner based on an instruction from the operation management apparatus.
- the total power consumption of the information processing device group in the computer room and the air conditioner can be reduced, so that the power saving operation of the data center can be realized.
- determination of workload allocation to the information processing device group does not depend on the output of the air conditioner, so there is no need to sequentially repeat workload allocation and power measurement, and power saving can be realized in a short time. .
- FIG. 1 It is a conceptual diagram which shows the COP calculation step output according to apparatus of a 3rd Example.
- the screen which the air-conditioning visualization analysis tool of the 4th example of the present invention provides is shown.
- the another screen which the air-conditioning visualization analysis tool of a 4th Example provides is shown.
- Air conditioner power table 500 Air conditioner specification table 500, 1500 Work load specification table 600 , 1900 Work load allocation table 900, 1600, 2400 Device-specific COP calculation step output result 1000, 1700 Device-related power consumption formula creation step output result 1100, 1800 Device evaluation index calculation step output result 2200 Air conditioner power table
- FIG. 1 is a block diagram of the first embodiment of the present invention.
- the information processing apparatus when the information processing apparatus is a server apparatus, the workload is allocated to the server apparatus so as to save the power of the entire computer room.
- this embodiment shows a case where there are two server apparatuses as information processing apparatuses, this number is arbitrary.
- storage devices and network devices may be mixed as information processing devices.
- the operation management apparatus 100 and the server apparatuses 110 a and 110 b are connected via the network 120.
- the workloads 111a to 111d correspond to an OS on a virtual machine, a server application such as a Web server and a DB server.
- the workload assignment agents 112a and 112b receive instructions from the operation management apparatus 100 and control the workloads 111a to 111d on the server apparatuses 110a and 110b.
- the workload assignment agent may be provided in the server apparatuses 110a and 110b, or may exist outside the server apparatuses 110a and 110b.
- the workload assignment agent is a virtual OS manager, which is provided in the server device.
- the workload assignment agent is a load balancer and exists outside the server device.
- the operation management device 100 includes arrangement information 200, an information processing device specification table 300, an air conditioner specification table 400, a work load specification table 600, a work load assignment table 500, a work load assignment unit 101, and a work load control unit 102. These elements may be arranged in one operation management apparatus 100 as shown in FIG. 1 or may be distributed in a plurality of apparatuses.
- Fig. 2 shows an example of the arrangement of each device in the computer room.
- the position information 200 set in the operation management apparatus 100 includes the apparatus IDs 221a and 221b of the server apparatuses 220a and 220b.
- the air conditioner IDs 211a and 211b of each air conditioner related to the operating environment of each server device, and the relative distance information 230a to 230c between each server device and each air conditioner 210a and 210b must be included.
- the arrangement information 200 may be in the form of a graphic as shown in FIG. 2 or in the form of a table.
- the information processing device specification table 300 shown in FIG. 3 relates to an information processing device in the computer room, for each device ID 310, device type 320, device power minimum value PDmini330, device power maximum value PDmaxi340, processing performance maximum value Wi350, assignable processing performance. 360, the resource utilization rate maximum value Xi 370, and the device power type PDi (x) 380 are held.
- the device ID 310 is an identifier for identifying the information processing devices 220a and 220b in the computer room.
- the device type 320 is information for identifying the type of the information processing device.
- the apparatus power minimum value PDmini 330 and the apparatus power maximum value PDmaxi 340 are the apparatus power and the maximum rated power in the idle state of the information processing apparatus, respectively.
- the apparatus power minimum value PDmini 330 and the apparatus power maximum value PDmaxi 340 are values unique to each information processing apparatus, and are specified in a catalog or the like.
- the device power type PDi (x) of the information processing device i assumes that the increase amount of the device power of the information processing device is proportional to the resource usage rate, and uses PDmini, PDmaxi, and the resource usage rate x ( It is expressed by Equation 10).
- the device power formula PDi (x) may be a constant for simplification or may be a formula different from (Equation 10) obtained by approximating the power consumption measurement result of the information processing device.
- the maximum processing performance value Wi350 in the server device is the maximum value of the arithmetic processing performance of the server device, and the performance value of a benchmark such as SPEC or the maximum value of the number of transactions per unit time of a certain application is used.
- the assignable processing performance 360 is assignable processing performance, and is equal to the processing performance maximum value Wi when no workload is assigned.
- the resource utilization maximum value Xi 370 in the server apparatus is an allocation maximum value of the CPU utilization ratio and the memory utilization ratio of the server apparatus, and is set by the administrator according to the utilization policy of the server apparatus.
- the resource utilization rate maximum value is set high so that more work load can be concentrated in the server device.
- the resource utilization maximum value is set low so that the work load is allocated with a margin for the resources of the server device.
- the air conditioner specification table 400 shown in FIG. 4 holds an air conditioner COP 420 for each air conditioner ID 410 regarding the air conditioners in the computer room.
- the air conditioner ID 410 is an identifier for identifying the air conditioners 210a and 210b in the computer room.
- the air conditioner COP420 is specified in a catalog or the like.
- the workload specification table 500 shown in FIG. 5 holds an allocation target device type 520, a workload amount 530, and an allocation policy 540 for each workload ID 510 regarding the workload.
- the work load ID 510 is information for identifying the work load executed in the computer room.
- the allocation target device type 520 is information indicating the type of information processing device to which the corresponding workload can be allocated.
- the workload amount 530 in the server device is an expected value of the arithmetic processing performance of the server device necessary for the operation of the corresponding workload, and is expressed using the same standard as the maximum processing performance value 350.
- the allocation policy 540 is information that is referred to when the workload allocation unit 101 determines the workload allocation. In FIG. 5, service priority and power saving priority are specified as the allocation policy. However, other allocation policies may be used, or the allocation policy may not be used.
- FIG. 6 is a table showing the relationship between each workload and the information processing apparatus to which the workload is assigned.
- the workload is identified using the workload ID 610, and the information processing apparatus that is the allocation destination is identified using the allocation apparatus ID 620.
- Fig. 7 shows the processing flow of the workload allocation unit.
- This processing flow includes a device-specific COP calculation step 701, a device power consumption formula creation step 702, a device evaluation index calculation step 703, an allocatable information processing device extraction step 704, and a workload allocation step 705.
- 200, an information processing device specification table 300, an air conditioner specification table 400, and a work load specification table 500 are input, and a work load allocation table 600 is output.
- the assignable information processing device extraction step 704 and the workload assignment step 705 are processes performed for all the workloads, and are performed for each workload.
- the device-specific COP calculation step 701 receives the arrangement information 200 and outputs an output result 900 shown in FIG.
- the output result 900 holds the distance 930 between each information processing device and the device-specific COP 940 of each information processing device for each air conditioner.
- the air conditioner is identified using an air conditioner ID 910, and the information processing apparatus is identified using a device ID 920.
- a distance Lji 930 between the air conditioner j and the information processing apparatus i is calculated by analyzing the arrangement information 200.
- the calculation of COPji 940 which is a device-specific COP, uses (Equation 3).
- the environmental constant A in (Equation 3) is set so that the average of COPji of the air conditioner j is COPj.
- the device-specific COP or the environmental constant A may be obtained by an experiment in a computer room or by a three-dimensional thermal fluid simulation.
- the air conditioner power for the device power can be obtained using a three-dimensional thermofluid simulation or a high-speed simulation engine that has extracted the main parameters based on the three-dimensional thermofluid simulation. is there.
- the device temperature can be determined from the device power.
- the information processing device specification table and COPji are input, and the device-related power consumption formula of each information processing device is output.
- FIG. 10 shows an output result 1000 of the device-related power consumption formula creation step 702.
- the output result 1000 holds a device-related power consumption formula 1020 for each device ID 1010.
- the device-related power consumption equation 1020 is calculated using (Equation 4), (Equation 5), and (Equation 6).
- the device evaluation index calculation step 703 receives the device-related power consumption formula 1000 and the information processing device specification table 300 as inputs, and outputs an evaluation index for each information processing device.
- FIG. 11 shows an output result 1100 of the apparatus evaluation index calculation step 703.
- the power saving performance evaluation index EEi 1120 and the service continuity evaluation index ESi 1130 are used as the evaluation indexes of each information processing apparatus.
- the power saving evaluation index EEi 1120 is a constant calculated based on (Equation 8), and indicates the performance per related power consumption at the maximum resource utilization rate of the information processing apparatus i.
- EEi is a power-saving evaluation index that considers air conditioner power and computer room layout information
- load assignment reduces the power consumption of the entire computer room by assigning work loads in order from the server device with the highest EEi value.
- the service continuity evaluation index ESi 1130 is a constant calculated based on (Equation 11), and indicates the free resource amount of the information processing apparatus at the maximum resource usage rate of the information processing apparatus i.
- By assigning the workload in order from the server device with the highest ESi value it is possible to determine the workload assignment with a margin in the information processing device.
- two evaluation indexes EEi and ESi are used for each information processing apparatus, but only the power saving evaluation index EEi may be used, or other evaluation indexes may be used.
- the assignable information processing device extraction step 704 extracts information processing devices to which a workload can be assigned, using the workload specification table 500 and the information processing device specification table 300 as inputs. Specifically, referring to the workload assignment target device type 520, the workload amount 530, the information processing device type 320, and the assignable processing performance 360, the device type 320 is equal to the assignment target device type 520, and the assignment is performed. An information processing apparatus whose possible processing performance 360 is larger than the workload 530 is extracted. In addition to the assignable processing performance 360, the information processing apparatus includes resources that need to be secured when assigning work loads.
- the information processing apparatus when the information processing apparatus is a server apparatus, this corresponds to the disk storage area, the number of OS or application licenses, and when the information processing apparatus is a storage apparatus, this corresponds to the number of connected clients.
- these resources do not appear explicitly in (Equation 8) and (Equation 11), they are factors for determining whether or not allocation is possible.
- the processing is shown on the assumption that these resources are sufficiently secured.
- the workload specification table 500 and the information processing device specification table 300 hold information on these resources, and information processing device extraction It may be used as a constraint condition.
- the workload allocation step 705 determines an information processing apparatus to which a workload is allocated according to the allocation policy from the information processing apparatus extracted in the allocatable information processing apparatus extraction step 704 and the output result 1100 of the apparatus evaluation index calculation step.
- FIG. 6 shows a workload allocation table 600 that is an output result of the workload allocation step 705.
- the workload allocation step 705 determines the information processing apparatus having the highest power-saving evaluation index EEi as the allocation-destination information processing apparatus when the allocation policy is power saving priority, and determines the service continuity when the allocation policy is service priority.
- the information processing apparatus with the highest evaluation index ESi is determined as the information processing apparatus to which the information is assigned.
- the value of the assignable processing performance 360 of the corresponding information processing apparatus is subtracted by the work load amount 530 of the work load.
- the above is the processing flow of the workload allocation unit.
- the workload from the device-specific COP calculation step 701 as shown in the processing flow of FIG. You may repeat until the allocation step 705 until a convergence condition is satisfy
- a device-specific COP is obtained from a three-dimensional thermofluid simulation
- a device power distribution which is input information of the three-dimensional thermofluid simulation
- a work load allocation step 705 is obtained in a work load allocation step 705
- a device-specific COP calculation step 701 is performed based on this.
- a series of processes of calculating another COP is repeated until the convergence condition is satisfied.
- the convergence condition in the determination 801 of the convergence condition includes that the difference from the previous calculation result of the device-specific COP is within a threshold, the workload allocation result is equal to the previous allocation result, and the like. Further, in the workload allocation step 705, there is a constraint that the apparatus temperature obtained using the three-dimensional thermal fluid simulation or the high-speed simulation engine does not exceed the operation upper limit temperature of the information processing apparatus or the temperature obtained by subtracting the margin from the operation upper limit temperature. May be added. In the work load allocating unit flow shown in FIG. 8, when the heat generation amount of the information processing apparatus is affected by the heat generated by the adjacent information processing apparatus, the apparatus power of the information processing apparatus is affected by the temperature in the computer room. It is also applicable to cases.
- the work load control unit 102 uses the work load allocation table 600 to control the work loads 111a to 111d of the information processing apparatuses 110a and 110b. This control differs depending on the type of information processing apparatus.
- the workload control unit 102 changes the settings of the virtual OS manager and the load distribution apparatus corresponding to the workload assignment agents 112a and 112b, and sends each server apparatus based on the workload assignment table 600. Allocate workload.
- the optimal allocation of the work load to the server device is determined in consideration of the air conditioner power, and the work load on the server device is controlled according to the assignment, thereby saving power in the computer room. Workload allocation can be realized.
- FIG. 12 is a diagram showing a second embodiment of the present invention.
- the information processing apparatus is a storage apparatus
- work load allocation is realized so as to save power in the entire computer room.
- this embodiment shows the case of two storage apparatuses as information processing apparatuses, this number is arbitrary. Further, server devices and network devices may be mixed as information processing devices.
- the operation management apparatus 1200, the storage apparatuses 1210a and 1210b, and the workload assignment agent 1220 are connected via the network 120.
- the workloads 1211a to 121d are data storage areas.
- the workload assignment agent 1220 receives an instruction from the operation management apparatus and controls the data storage area of the storage apparatus.
- the storage device workload assignment agent 1220 is a storage management manager, and usually exists outside the storage device.
- the 12 includes an arrangement information 1300, an information processing device specification table 1400, an air conditioner specification table 400, a work load specification table 1500, a work load assignment table 1900, a work load assignment unit 1201, and a work load control unit 1202. Is provided.
- the arrangement information 1300, the information processing device specification table 1400, the work load specification table 1500, the work load assignment unit 1201, and the work load control unit 1202, which are differences from the first embodiment, will be described.
- Air conditioner IDs 1311a and 1311b are individually assigned to the air conditioners 1310a and 1310b, and device IDs 1321a and 1321b are individually assigned to the storage apparatuses 1320a and 1320b.
- the arrangement information includes relative distance information 1330a to 1330c between the air conditioners 1310a and 1310b and the storage apparatuses 1320a and 1320b. If the relative distance between the air conditioner and each storage device is known, the arrangement information may be in the form of a graphic shown in FIG. 13 or in the form of a table.
- the information processing apparatus specification table 1400 illustrated in FIG. 14 relates to a storage apparatus in a computer room, for each apparatus ID 1410, a device type 1420, a device power minimum value PDmini 1430, a device power maximum value PDmaxi 1440, a processing performance maximum value Wi 1450, and an assignable processing performance 1460.
- Resource utilization rate maximum value Xi 1470 and device power expression 1480 are held.
- the processing performance maximum value Wi 1450 is the maximum value of the data storage area of the storage apparatus and is specified in a catalog or the like.
- the resource utilization rate maximum value Xi 1470 is the maximum value of the utilization rate of the data storage area of the storage apparatus, and is referred to by the workload assignment unit 1201 when assigning the workload.
- the work load specification table 1500 illustrated in FIG. 15 holds an assignment target device type 1520 and a work load amount 1530 for each work load ID 1510 for the work load.
- the work load amount 1530 in the storage device is the size of the data storage area required by the corresponding work load.
- the allocation policy may be defined as in the workload specification table 500 of the first embodiment.
- the processing flow of the workload assignment unit 1201 in the present embodiment is equivalent to the work flow assignment unit processing flow 700 of the first embodiment.
- the device-specific COP calculation step 701 receives the arrangement information 1300 and the air conditioner specification table 400, and outputs the device-specific COP for each air conditioner shown in the output result 1600.
- the information processing device specification table 1400 and the output result 1600 of the device-specific COP calculation step are input, and the device-related power consumption formula of each information processing device shown in the output result 1700 is obtained. Output.
- the device power of the storage device is constant regardless of the resource utilization rate, and the device power consumption formula Pi is set with the device power formula PDi (x) of the storage device being constant at the device power maximum value PDmaxi. (X) was determined.
- the storage device adopts MAID (Massive Arrays of Inactive Disks) technology that performs power control according to the storage area used
- the device power type PDi ( x) and the device-related power consumption formula Pi (x) may be obtained.
- the device evaluation index calculation step 703 in this embodiment the device-related power consumption formula 1700 and the information processing device specification table 1400 are input, and the evaluation index of each storage device shown in the output result 1800 is output. In this embodiment, only the power saving evaluation index EEi is output. However, other evaluation indexes may be output as described in the first embodiment.
- the assignable information processing device extraction step 704 and the workload assignment step 705 in the present embodiment receive the evaluation index 1800, the workload specification table 1500, and the information processing device specification table 1400 of each information processing device as inputs, and work shown in FIG.
- the load allocation table 1900 is output.
- the evaluation index 1800 of each information processing device the storage device 1320a with the device ID 3 has a higher power saving evaluation index EEi than the storage device 1320b with the device ID 4. Therefore, the workload allocation table 1900 has a work related to the storage device.
- the load is collected in the storage device 1320a. This indicates that power including the air conditioner power in the computer room can be saved by concentrating work loads on the storage device 1320a.
- the work load control unit 1202 controls the work loads of the information processing apparatuses 1210a and 1210b using the work load assignment table 1900. This control differs depending on the type of information processing apparatus.
- the workload control unit changes the setting of the storage management manager corresponding to the workload allocation agent 1220 and stores data that is the workload of each storage apparatus based on the workload allocation table 1900. Allocate space.
- FIG. 20 is a diagram showing a third embodiment of the present invention.
- the air conditioner power necessary for cooling the information processing devices 2010a to 2010d in the computer room is obtained, and the air conditioners 2020a and 2020b are controlled accordingly, thereby reducing the power of the entire computer room.
- the present embodiment shows the case of four information processing devices and two air conditioners, this number is arbitrary.
- the present embodiment may be performed in parallel with the first embodiment or the second embodiment, or may be performed alone.
- the operation management apparatus 2000, the information processing apparatuses 2010a to 2010d, and the air conditioners 2020a and 2020b are connected by a network 120.
- the information processing devices 2010a to 2010d in the computer room include power measurement agents 2011a to 2011d.
- the power measurement agents 2011a to 2011d measure the apparatus power of the information processing apparatus and notify the operation management apparatus 2000 of the result.
- the air conditioners 2020a and 2020b include output control agents 2021a and 2021b.
- the output control agents 2021a and 2021b control the outputs of the air conditioners 2020a and 2020b in accordance with a request from the operation management apparatus 2000.
- the operation management apparatus 2000 includes arrangement information 2100, an air conditioner specification table 400, an air conditioner power table 2200, an air conditioner power calculation unit 2001, and an air conditioner control unit 2002.
- Air conditioner IDs 2111a and 2111b are individually assigned to the air conditioners 2110a and 2110b, and apparatus IDs 2121a to 2121d are individually assigned to the information processing apparatuses 2120a to 2120d.
- the arrangement information 2100 includes relative distance information 2130a to 2130d between the air conditioners 2110a and 2110b and the information processing apparatuses 2120a to 2120d.
- the arrangement information may be in the form of a graphic as shown in FIG. 21 or in the form of a table as long as the relative distance between the air conditioner and each information processing device is known.
- the air conditioner power 2250 is calculated from the device power measured value 2230 and the air conditioner power formula 2240 for each device ID 2220 of the information processing device.
- the air conditioner power table 2200 is created by the air conditioner power calculation unit 2001, and the air conditioner control unit 2002 is referred to in order to control the outputs of the air conditioners 2110a and 2110b.
- FIG. 23 shows a processing flow of the air conditioner power calculation unit 2001.
- This processing flow includes a device-specific COP calculation step 2301, a device power acquisition step 2302 and an air conditioner power calculation step 2303.
- the arrangement information 2100, the air conditioner specification table 400, and the device power measurement value 2230 of the information processing device are input.
- the air conditioner power table 2200 is output.
- the device-specific COP calculation step 2301 is equivalent to the device-specific COP calculation step 701 of the first embodiment.
- the device-specific COP calculation step 2301 receives the arrangement information 2100 and the air conditioner specification table 400 and outputs the device-specific COP.
- FIG. 24 shows an output result 2400 of the device-specific COP calculation step 2301.
- Device power acquisition step 2302 acquires the device power measurement value 2230 of the information processing devices 2010a to 2010d via the device power measurement agents 2011a to 2011d.
- a value calculated using the device power maximum value or the device power formula PDi (x) may be used instead of the device power measurement value.
- Air conditioner power table 2200 shown in FIG. 21 is an output result of air conditioner power calculation step 2303.
- the air conditioner power PCj2250 of each air conditioner is calculated using (Equation 9). Since the air conditioner power PCj2250 of each air conditioner is an optimum value of the air conditioner power, the air conditioner power in the computer room can be reduced by controlling the outputs of the air conditioners 2010a and 2010b based on this value.
- the air conditioner control unit 2002 refers to the air conditioner power table 2200 and controls the outputs of the air conditioners 2020a and 2020b via the output control agents 2021a and 2021b.
- the present embodiment can be implemented in parallel with the assignment of the workload to the information processing apparatus shown in the first embodiment or the second embodiment.
- the allocation of the work load to the information processing apparatus is optimized, and the total value of the computer room apparatus power and the air conditioner power is reduced. It becomes possible.
- Example 1 or Example 2 does not depend on the result of this example, it is not necessary to sequentially repeat the assignment of workload and the power measurement, and the time until realizing power saving can be shortened.
- FIG. 25 shows an example of a screen displayed by the air conditioning visualization analysis tool provided in the operation management apparatus 100 of FIG.
- the air conditioning visualization analysis tool includes a computer room layout display screen 2510, an apparatus / air conditioner selection screen 2520, a detailed information screen 2530, a computer room information DB selection button 2540, an air conditioner display item selection box 2550, an apparatus in the entire window 2500.
- a display item selection box 2560 is displayed, and air conditioning analysis information is provided to the administrator.
- the computer room layout display screen 2510 includes a computer room layout outline 2511, a device object 2512, and air conditioner objects 2513 and 2514, and provides air conditioner and device arrangement information to the administrator.
- the device / air conditioner selection screen 2520 displays a list of devices and air conditioners in the computer room, and provides the administrator with devices and air conditioner selection means to be displayed on the detailed information screen 2530.
- the air conditioning visualization analysis tool 2500 highlights the corresponding object 2514 on the computer room layout display screen 2510.
- detailed information screen 2530 displays detailed information of the corresponding object.
- “air conditioner 2” 2521 on the device / air conditioner selection screen 2520 is highlighted and the details of the corresponding object are displayed on the detailed information screen 2530. Display information.
- the display items of the detailed information screen 2530 are the device name, device type, device power, device-specific COP, and COP (average value of device-specific COP) related to the selected “air conditioner 2” 2521. However, other items may be included.
- the DB selection button 2540 is a button for calling up a screen for selecting information of a computer room to be subjected to air conditioning visualization analysis.
- the button format is used, but other formats may be used.
- the air conditioner display item selection box 2550 is a box for the administrator to specify display items related to the air conditioner object on the computer room layout display screen 2510.
- a device display function selection box 2560 is a box for designating display items related to a device object.
- the item (e) is designated in the air conditioner display item selection box 2550, and as a result, the COP value of each air conditioner is displayed in the air conditioner objects 2513 and 2514 on the computer room layout display screen 2510. Indicates the displayed state.
- the item (h) is specified in the device display function selection box 2560, and as a result, the value of the resource utilization rate of each storage device is shown inside the device object 2512 on the computer room layout display screen 2510. .
- the administrator selects “device 4” 2610 on the device / air conditioner selection screen 2520.
- the storage device object 2620 is highlighted on the computer room layout display screen, and the object is displayed on the detailed information screen.
- the detailed information is displayed.
- the display items of the detailed display of the storage device are the device name, device type, device power, resource utilization rate, power saving evaluation index, device-specific COP for each air conditioner, and device-specific COP for each air conditioner. Average value.
- the administrator can realize the computer room operation with low operation cost. For example, when “COP by device (device 1)” is displayed on the air conditioner object on the computer room layout display screen 2510, it is possible to confirm the influence of the device power increase of the device 1 on the air conditioner. That is, the administrator can confirm which air conditioner is more effective to control in accordance with the increase or decrease of the operating status of the device 1. Control by the administrator using the result can reduce the power of the air conditioner.
- “power saving evaluation index” is displayed on the device object on the computer room layout display screen 2510, when adding a new work load, to which device the work load is assigned, the power efficiency is high. Can be confirmed. Control by the administrator using the result can suppress an increase in power of the entire computer room.
- “device-specific COP (air conditioner 1)” can be displayed on the device object on the computer room layout display screen 2510. For example, it can be confirmed which work load is assigned to which device can suppress the increase in the output of the air conditioner 1 to a minimum.
- the air conditioner in the computer room can be used as much as possible, and the timing of newly introducing the air conditioner can be delayed.
- the present invention can be applied to integrated power-saving operation management in an information processing system composed of a group of information processing devices such as servers, storage, and networks and a cooling facility, and is particularly suitable for a data center.
- the present invention can be applied to a wide range of applications as a system construction tool, an energy saving diagnostic tool, an operation monitoring tool, and an auxiliary tool for an operation manager or facility manager.
- the present invention is mainly intended for the information processing apparatus group, but the present invention can be applied to any apparatus that consumes electric power or energy and equipment therefor.
- the present invention can be applied to the operation and control of electric devices, mechanical devices, power devices, heat devices, and the like.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- Human Computer Interaction (AREA)
- Air Conditioning Control Device (AREA)
- Power Sources (AREA)
Abstract
Description
空調機の情報処理装置に関する空調機動作係数は、空調機が特定の情報処理装置の発熱を交換する際のCOPであり、ここでは装置別COPと呼ぶ。空調機jの情報処理装置iに関する装置別COPであるCOPjiは、情報処理装置iの発熱量PDi、空調機jの情報処理装置iに関する空調機電力PCjiを用いて(数2)で表せる。
一般的に熱は拡散するため、装置別COPは空調機と情報処理装置の間の距離が長くなるほど低下する。熱の拡散は気流、風量、温度に依存することを考慮すると、装置別COPは、三次元熱流体シミュレーションの結果等から求める必要がある。ただし、他の方法で装置別COPを求めても良い。一例として、問題を簡易化して、熱が2次元に広がり、装置別COPが空調機と情報処理装置の距離の2乗に反比例すると仮定すると、空調機jの情報処理装置iに関する装置別COPであるCOPjiは、空調機jと情報処理装置iの間の距離Lji、空調機jに固有の空調機動作係数COPj、並びに環境定数Aを用いて(数3)で表せる。空調機動作係数COPjは空調機jの仕様より、距離Ljiは計算機室における空調機と情報処理装置群の位置から成る配置情報より求めることができる。
装置別COPによって、ある情報処理装置に関する空調機電力を計算可能である。1台の空調機jが情報処理装置iの発熱を処理する場合、情報処理装置の発熱量は装置電力と等しいと仮定し、空調機jの情報処理装置iに関する空調機電力PCjiは、情報処理装置iの装置電力PDiとCOPjiを用いて(数4)で表せる。また、複数の空調機が情報処理装置iの発熱を処理する場合、情報処理装置iに関する空調機電力PCiは、空調機が処理する情報処理装置の発熱量が各空調機のCOPji比に比例すると仮定し、空調機台数N、PDi、COPjiを用いて(数5)で表せる。
ここでPDi(x)、PCi(x)は、情報処理装置のリソース利用率xの関数である。
情報処理装置がサーバ装置の場合、リソース利用率は主にCPU利用率やメモリ利用率である。情報処理装置がストレージ装置の場合、リソース利用率は主にディスク領域利用率である。Pi(x)は、装置電力と空調機電力の合計であり、この空調機電力は計算機室内の空調機と情報処理装置群の位置情報が考慮されている。すなわち、Pi(x)を用いることで、計算機室内の位置情報を意識することなく、計算機室内の位置情報を考慮した消費電力の評価が可能である。
ここで、EEiは定数であり、値が大きいほど情報処理装置iの性能対消費電力性能が優れる事を示す。ここで示す消費電力とは情報処理装置iに関する装置電力と空調機電力の合計値である。すなわち、省電力性評価指標EEiが大きい情報処理装置iに優先して作業負荷を割り当てることで、計算機室全体の消費電力が最小となる作業負荷の割当が可能である。この方針で作業負荷の割当を行うと、省電力性評価指標が大きい情報処理装置のリソース利用率は最大リソース利用率に近づき、省電力性評価指標が小さい情報処理装置のリソース利用率がゼロに近づく二極化現象が現れる。省電力性評価指標EEiでは、この二極化現象を考慮して、情報処理装置のリソース利用率がリソース利用率最大値の場合の省電力性を評価する。
110a~110b,1210a~1210b,2010a~2010d 情報処理装置
200,1300,2100 配置情報
300,1400 情報処理装置仕様表
400 空調機仕様表
500,1500 作業負荷仕様表
600,1900 作業負荷割当表
900,1600,2400 装置別COP計算ステップ出力結果
1000,1700 装置関連消費電力式作成ステップ出力結果
1100,1800 装置評価指標計算ステップ出力結果
2200 空調機電力表
サーバ装置における処理性能最大値Wi350は、サーバ装置の演算処理性能の最大値であり、SPEC等のベンチマークの性能値もしくは、あるアプリケーションの単位時間あたりのトランザクション数の最大値を用いる。割当可能処理性能360は、割当可能な処理性能であり、作業負荷が割り当てられていない場合は処理性能最大値Wiと等しい。また、サーバ装置におけるリソース利用率最大値Xi370は、サーバ装置のCPU利用率やメモリ利用率の割当最大値であり、管理者がサーバ装置の利用ポリシに従い設定する。例えば、省電力性を優先する場合は、より多くの作業負荷をサーバ装置に集約できるよう、リソース利用率最大値を高く設定する。また、サービス継続性を優先する場合は、サーバ装置のリソースに余裕を持たせて作業負荷を割り当てるよう、リソース利用率最大値を低く設定する。
割当可能情報処理装置抽出ステップ704は、作業負荷仕様表500、情報処理装置仕様表300を入力として、作業負荷を割当可能な情報処理装置を抽出する。具体的には、作業負荷の割当対象装置種別520、作業負荷量530および情報処理装置の装置種別320、割当可能処理性能360を参照して、装置種別320が割当対象装置種別520と等しく、割当可能処理性能360が作業負荷量530より大きい情報処理装置を抽出する。情報処理装置は、割当可能処理性能360の他に、作業負荷の割当の際に確保が必要なリソースを含む。例えば、情報処理装置がサーバ装置の場合、ディスク格納領域やOSやアプリケーションのライセンス数等が該当し、情報処理装置がストレージ装置の場合接続クライアント数等が該当する。これらリソースは、(数8)、(数11)には明示的に現れていないが、割当可否の判断の因子となる。本実施例では、これらリソースが十分確保されている仮定での処理を示しているが、作業負荷仕様表500や情報処理装置仕様表300がこれらリソースの情報を保持し、情報処理装置の抽出の際に制約条件として用いても良い。
Claims (11)
- 計算機室の複数の情報処理装置および空調機の配置情報、各空調器の仕様情報、および各情報処理装置の仕様情報を予め格納した運用管理装置により前記複数の情報処理装置への作業負荷配分を決定して省電力運用を実現する情報処理のシステム運用管理方法であって、
前記配置情報から抽出した各情報処理装置と冷却に寄与する各空調機の間の相対距離および各空調機の仕様に基づき各情報処理装置に関する装置別空調機動作係数を求め、
各情報処理装置の消費電力と各情報処理装置あたりの空調機消費電力の合計を表す装置関連消費電力式を、前記装置別空調機動作係数を用いて作成し、
作成した前記各情報処理装置の装置関連消費電力式と各情報処理装置の性能から各情報処理装置の単位消費電力あたりの性能を示す省電力性評価指数を算出し、
省電力性評価指数が低い情報処理装置よりも省電力性評価指数が高い情報処理装置を優先して作業負荷を割り当てることで作業負荷配分を決定することを特徴とする情報処理システムの運用管理方法。 - 前記情報処理装置はサーバ装置であり、前記省電力性評価指標は、該サーバ装置の消費電力と該サーバ装置あたりの空調機消費電力の合計消費電力と該サーバ装置の処理性能を用いて算出した、単位消費電力あたりの処理性能であることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記情報処理装置はストレージ装置であり、前記作業負荷はストレージ装置内に設けられたデータ格納領域であり、前記省電力性評価指標は、該ストレージ装置の消費電力と該ストレージ装置あたりの空調機消費電力の合計消費電力と該ストレージ装置のデータ格納領域を用いて算出した、単位消費電力あたりのデータ格納領域であることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 作業負荷が配分された各情報処理装置の装置電力測定値を取得する手順、
前記求められた装置別空調機動作係数と該取得した各情報処理装置の装置電力測定値を用い、空調機電力式により前記各空調機の電力計算値を算出する手順、および
算出した電力計算値に各空調機の運転電力を制御する手順を更に有することを特徴とする請求項1記載の情報処理システムの運用管理方法。 - 前記作業負荷配分では、割り当てを決定しようとする作業負荷が割り当て可能な情報処理装置の中から前記省電力性評価指数が最も大きい情報処理装置を選んで割り当て、もって省電力性評価指標の高い情報処理装置から順に作業負荷を割当てることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記装置別空調機動作係数は、空調機に固有の動作係数に、対応する情報処理装置と該空調機との相対距離の2乗に反比例した係数を掛けて算出することを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記各情報処理装置の装置関連消費電力式は、該情報処理装置のリソース利用率の関数であることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記省電力性評価指標は、前記装置関連消費電力式に管理者の定めるリソース利用率最大値を代入して算出した単位電力あたりの処理性能であることを特徴とする請求項7記載の情報処理システムの運用管理方法。
- 前記複数の情報処理装置に配分すべき作業負荷ごとに、省電力を優先した配分か否かを示す割り当てポリシが付され、省電力を優先した配分である場合にかぎり、割り当て可能な情報処理装置のうちで前記省電力性評価指標が最も大きな情報処理装置に当該作業負荷を割り当てることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記省電力性評価指標の算出に加えて、前記各情報処理装置ごとに、その最大リソース利用率における空リソース量を示すサービス継続性評価指数を算出し、前記各情報処理装置に配分すべき作業負荷ごとに、省電力を優先した配分かサービス継続性を優先した配分かを示す割り当てポリシが付され、省電力を優先した配分である場合には割り当て可能な情報処理装置のうちで前記省電力性評価指標が最も大きな情報処理装置に当該作業負荷を割り当て、サービス継続性を優先した配分の場合は前記サービス継続性評価指数が最も大きな情報処理装置に当該作業負荷を割り当てることを特徴とする請求項1記載の情報処理システムの運用管理方法。
- 前記各情報処理装置のサービス継続性評価指数は、当該情報処理装置の最大リソース利用率における空リソース量から算出することを特徴とする請求項10記載の情報処理システムの運用管理方法。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US13/003,687 US8145927B2 (en) | 2008-09-17 | 2009-03-09 | Operation management method of information processing system |
CN2009801278238A CN102099791B (zh) | 2008-09-17 | 2009-03-09 | 信息处理系统的运用管理方法 |
EP09814346.4A EP2330505B1 (en) | 2008-09-17 | 2009-03-09 | Operation management method of information processing system |
JP2010529664A JP4751962B2 (ja) | 2008-09-17 | 2009-03-09 | 情報処理システムの運用管理方法 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2008237325 | 2008-09-17 | ||
JP2008-237325 | 2008-09-17 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010032501A1 true WO2010032501A1 (ja) | 2010-03-25 |
Family
ID=42039348
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2009/054446 WO2010032501A1 (ja) | 2008-09-17 | 2009-03-09 | 情報処理システムの運用管理方法 |
Country Status (5)
Country | Link |
---|---|
US (1) | US8145927B2 (ja) |
EP (1) | EP2330505B1 (ja) |
JP (2) | JP4751962B2 (ja) |
CN (1) | CN102099791B (ja) |
WO (1) | WO2010032501A1 (ja) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011227635A (ja) * | 2010-04-19 | 2011-11-10 | Hitachi Ltd | 情報処理システム及び情報処理方法 |
JP2012073784A (ja) * | 2010-09-28 | 2012-04-12 | Fujitsu Ltd | 計算処理システム、そのジョブ分散配置方法及びジョブ分散配置プログラム |
JP2012118727A (ja) * | 2010-11-30 | 2012-06-21 | Okamura Corp | エリア利用状況解析システム及びエリア利用状況解析方法並びにそのプログラム |
JP2013073402A (ja) * | 2011-09-27 | 2013-04-22 | Fujitsu Ltd | 電力制御装置、電力制御方法、および電力制御プログラム |
US8473108B2 (en) | 2010-11-09 | 2013-06-25 | Hitachi, Ltd. | Information apparatus and method of optimizing cooling efficiency of air conditioner |
US8761955B2 (en) | 2011-04-27 | 2014-06-24 | Hitachi, Ltd. | Management computer, computer system including the same, and method for providing allocating plan for it equipment |
JP2017151617A (ja) * | 2016-02-23 | 2017-08-31 | 日本電信電話株式会社 | シミュレーション装置、シミュレーション方法及びプログラム |
US11137817B2 (en) | 2016-03-30 | 2021-10-05 | Fujitsu Limited | Recording medium, arrangement search method, and arrangement searching apparatus |
Families Citing this family (30)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP4958883B2 (ja) * | 2008-10-29 | 2012-06-20 | 株式会社日立製作所 | 管理サーバ装置によるストレージ装置及び空調装置の制御方法及びストレージシステム |
JP5098978B2 (ja) * | 2008-12-02 | 2012-12-12 | 富士通株式会社 | 消費電力削減支援プログラム、情報処理装置、および消費電力削減支援方法 |
US8214829B2 (en) | 2009-01-15 | 2012-07-03 | International Business Machines Corporation | Techniques for placing applications in heterogeneous virtualized systems while minimizing power and migration cost |
US8510582B2 (en) * | 2010-07-21 | 2013-08-13 | Advanced Micro Devices, Inc. | Managing current and power in a computing system |
CN102446197B (zh) * | 2010-09-30 | 2015-11-25 | 国际商业机器公司 | 一种数据转换方法和数据转换器 |
US8694279B1 (en) | 2010-12-30 | 2014-04-08 | Exaflop Llc | Data center thermal monitoring |
US20120233236A1 (en) * | 2011-03-07 | 2012-09-13 | Min-Shu Chen | Cloud-based system for serving service request of embedded device by cloud computing and related cloud-based processing method thereof |
US9195510B2 (en) | 2011-04-04 | 2015-11-24 | Dell Products L.P. | Information handling system application decentralized workload management |
US8793686B2 (en) * | 2011-06-08 | 2014-07-29 | Microsoft Corporation | Operating system decoupled heterogeneous computing |
JP5776427B2 (ja) * | 2011-08-04 | 2015-09-09 | 富士通株式会社 | 情報処理システム及び情報処理方法 |
JP5724753B2 (ja) * | 2011-08-25 | 2015-05-27 | 富士通株式会社 | 吸着式ヒートポンプの制御方法、情報処理システム及び制御装置 |
JP5568535B2 (ja) | 2011-09-28 | 2014-08-06 | 株式会社日立製作所 | データセンタの負荷割当て方法及び情報処理システム |
US9229786B2 (en) * | 2011-10-25 | 2016-01-05 | International Business Machines Corporation | Provisioning aggregate computational workloads and air conditioning unit configurations to optimize utility of air conditioning units and processing resources within a data center |
JP5785050B2 (ja) | 2011-10-27 | 2015-09-24 | 株式会社日立製作所 | 情報処理システム、その省電力制御方法、及び装置 |
EP2590045A1 (en) * | 2011-11-03 | 2013-05-08 | Danfoss A/S | A method for setting parameters in a system, in particular a heating or cooling system, device to change parameters, and heating or cooling system |
US8862909B2 (en) | 2011-12-02 | 2014-10-14 | Advanced Micro Devices, Inc. | System and method for determining a power estimate for an I/O controller based on monitored activity levels and adjusting power limit of processing units by comparing the power estimate with an assigned power limit for the I/O controller |
US8924758B2 (en) | 2011-12-13 | 2014-12-30 | Advanced Micro Devices, Inc. | Method for SOC performance and power optimization |
JP5736302B2 (ja) | 2011-12-15 | 2015-06-17 | 株式会社日立製作所 | 情報処理システム、情報処理システムの運用管理方法、およびデータセンタ |
JP5801732B2 (ja) * | 2012-01-24 | 2015-10-28 | 株式会社日立製作所 | 情報処理システムの運用管理方法 |
WO2013145273A1 (ja) * | 2012-03-30 | 2013-10-03 | 富士通株式会社 | 情報処理装置、制御方法、及びプログラム |
US9207744B2 (en) * | 2012-10-18 | 2015-12-08 | Huawei Technologies Co., Ltd. | Method and apparatus for adjusting device power consumption |
KR20140079274A (ko) * | 2012-12-18 | 2014-06-26 | 삼성전자주식회사 | 홈 네트워크 시스템에서 에너지 소비를 관리하는 방법 및 장치 |
CN104321593B (zh) * | 2013-02-20 | 2017-08-15 | 松下电器(美国)知识产权公司 | 便携信息终端及其控制方法 |
US10114719B2 (en) * | 2013-02-21 | 2018-10-30 | International Business Machines Corporation | Estimating power usage in a computing environment |
JP5969939B2 (ja) | 2013-02-28 | 2016-08-17 | 株式会社日立製作所 | データセンタの空調制御装置 |
JP5988505B2 (ja) * | 2013-11-26 | 2016-09-07 | 日本電信電話株式会社 | 仮想リソース管理装置、選択方法及び選択プログラム |
US10423217B1 (en) * | 2017-07-14 | 2019-09-24 | Cisco Technology, Inc. | Dynamic power capping of multi-server nodes in a chassis based on real-time resource utilization |
US11073888B2 (en) * | 2019-05-31 | 2021-07-27 | Advanced Micro Devices, Inc. | Platform power manager for rack level power and thermal constraints |
CN110399216B (zh) * | 2019-06-27 | 2021-10-15 | 苏州浪潮智能科技有限公司 | 一种整机箱功耗的分配方法、系统、装置及可读存储介质 |
CN111126786B (zh) * | 2019-11-29 | 2023-08-22 | 青岛海尔科技有限公司 | 用于设备调度的方法及装置、服务器 |
Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2004126968A (ja) | 2002-10-03 | 2004-04-22 | Fujitsu Ltd | 並列計算機のジョブスケジューリング装置 |
JP2005531047A (ja) * | 2002-03-18 | 2005-10-13 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 複数コンピュータ・サーバの電力消費を管理する方法 |
JP2005312142A (ja) * | 2004-04-20 | 2005-11-04 | Hitachi Ltd | 消費電力量管理システム |
JP2005321162A (ja) * | 2004-05-11 | 2005-11-17 | Hitachi Ltd | 計算機システム用空調設備及びその温度制御方法 |
JP2006064253A (ja) * | 2004-08-25 | 2006-03-09 | Ntt Power & Building Facilities Inc | 空調機監視システム、および空調機監視方法 |
US20060259793A1 (en) | 2005-05-16 | 2006-11-16 | Justin Moore | Power distribution among servers |
US20060259621A1 (en) | 2005-05-16 | 2006-11-16 | Parthasarathy Ranganathan | Historical data based workload allocation |
JP2007179437A (ja) | 2005-12-28 | 2007-07-12 | Fujitsu Ltd | 管理システム、管理プログラムおよび管理方法 |
JP2007536657A (ja) * | 2004-05-08 | 2007-12-13 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 仮想マシン・コンピュータ・プログラムの動的マイグレーション |
Family Cites Families (17)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11296488A (ja) | 1998-04-09 | 1999-10-29 | Hitachi Ltd | 電子機器 |
WO2003083693A1 (fr) | 2002-04-03 | 2003-10-09 | Fujitsu Limited | Planificateur de taches dans un systeme de traitement distribue |
US20030193777A1 (en) * | 2002-04-16 | 2003-10-16 | Friedrich Richard J. | Data center energy management system |
JP2004240669A (ja) * | 2003-02-05 | 2004-08-26 | Sharp Corp | ジョブスケジューラおよびマルチプロセッサシステム |
JP3896352B2 (ja) | 2003-08-08 | 2007-03-22 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 分散コンピューティングシステム |
JP2006062453A (ja) | 2004-08-25 | 2006-03-09 | Toyota Motor Corp | 車両用制御装置 |
US20060112286A1 (en) * | 2004-11-23 | 2006-05-25 | Whalley Ian N | Method for dynamically reprovisioning applications and other server resources in a computer center in response to power and heat dissipation requirements |
JP2006285317A (ja) | 2005-03-31 | 2006-10-19 | Tokyo Electric Power Co Inc:The | 負荷判定システム、負荷分散システムおよび異常検出システム |
US7596476B2 (en) * | 2005-05-02 | 2009-09-29 | American Power Conversion Corporation | Methods and systems for managing facility power and cooling |
US7885795B2 (en) * | 2005-05-02 | 2011-02-08 | American Power Conversion Corporation | Methods and systems for managing facility power and cooling |
JP4476876B2 (ja) | 2005-06-10 | 2010-06-09 | 三菱電機株式会社 | 並列計算装置 |
CN200969053Y (zh) * | 2006-06-19 | 2007-10-31 | 侨威科技股份有限公司 | 具有智能型冷却控制功能的电源供应器 |
CN2917201Y (zh) * | 2006-06-20 | 2007-06-27 | 华为技术有限公司 | 一种电路板散热结构 |
US7739388B2 (en) * | 2007-05-30 | 2010-06-15 | International Business Machines Corporation | Method and system for managing data center power usage based on service commitments |
US8635625B2 (en) * | 2008-04-04 | 2014-01-21 | International Business Machines Corporation | Power-aware workload allocation in performance-managed computing environments |
JP4724730B2 (ja) * | 2008-04-09 | 2011-07-13 | 株式会社日立製作所 | 情報処理システムの運用管理方法、運用管理プログラム、および運用管理装置、ならびに情報処理システム |
US7472558B1 (en) * | 2008-04-15 | 2009-01-06 | International Business Machines (Ibm) Corporation | Method of determining optimal air conditioner control |
-
2009
- 2009-03-09 WO PCT/JP2009/054446 patent/WO2010032501A1/ja active Application Filing
- 2009-03-09 US US13/003,687 patent/US8145927B2/en not_active Expired - Fee Related
- 2009-03-09 JP JP2010529664A patent/JP4751962B2/ja not_active Expired - Fee Related
- 2009-03-09 EP EP09814346.4A patent/EP2330505B1/en not_active Not-in-force
- 2009-03-09 CN CN2009801278238A patent/CN102099791B/zh active Active
-
2010
- 2010-09-29 JP JP2010218354A patent/JP2011034578A/ja active Pending
Patent Citations (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005531047A (ja) * | 2002-03-18 | 2005-10-13 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 複数コンピュータ・サーバの電力消費を管理する方法 |
JP2004126968A (ja) | 2002-10-03 | 2004-04-22 | Fujitsu Ltd | 並列計算機のジョブスケジューリング装置 |
JP2005312142A (ja) * | 2004-04-20 | 2005-11-04 | Hitachi Ltd | 消費電力量管理システム |
JP2007536657A (ja) * | 2004-05-08 | 2007-12-13 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 仮想マシン・コンピュータ・プログラムの動的マイグレーション |
JP2005321162A (ja) * | 2004-05-11 | 2005-11-17 | Hitachi Ltd | 計算機システム用空調設備及びその温度制御方法 |
JP2006064253A (ja) * | 2004-08-25 | 2006-03-09 | Ntt Power & Building Facilities Inc | 空調機監視システム、および空調機監視方法 |
US20060259793A1 (en) | 2005-05-16 | 2006-11-16 | Justin Moore | Power distribution among servers |
US20060259621A1 (en) | 2005-05-16 | 2006-11-16 | Parthasarathy Ranganathan | Historical data based workload allocation |
JP2007179437A (ja) | 2005-12-28 | 2007-07-12 | Fujitsu Ltd | 管理システム、管理プログラムおよび管理方法 |
Non-Patent Citations (1)
Title |
---|
See also references of EP2330505A4 |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2011227635A (ja) * | 2010-04-19 | 2011-11-10 | Hitachi Ltd | 情報処理システム及び情報処理方法 |
JP2012073784A (ja) * | 2010-09-28 | 2012-04-12 | Fujitsu Ltd | 計算処理システム、そのジョブ分散配置方法及びジョブ分散配置プログラム |
US8473108B2 (en) | 2010-11-09 | 2013-06-25 | Hitachi, Ltd. | Information apparatus and method of optimizing cooling efficiency of air conditioner |
JP2012118727A (ja) * | 2010-11-30 | 2012-06-21 | Okamura Corp | エリア利用状況解析システム及びエリア利用状況解析方法並びにそのプログラム |
US8761955B2 (en) | 2011-04-27 | 2014-06-24 | Hitachi, Ltd. | Management computer, computer system including the same, and method for providing allocating plan for it equipment |
JP2013073402A (ja) * | 2011-09-27 | 2013-04-22 | Fujitsu Ltd | 電力制御装置、電力制御方法、および電力制御プログラム |
JP2017151617A (ja) * | 2016-02-23 | 2017-08-31 | 日本電信電話株式会社 | シミュレーション装置、シミュレーション方法及びプログラム |
US11137817B2 (en) | 2016-03-30 | 2021-10-05 | Fujitsu Limited | Recording medium, arrangement search method, and arrangement searching apparatus |
Also Published As
Publication number | Publication date |
---|---|
EP2330505A1 (en) | 2011-06-08 |
US8145927B2 (en) | 2012-03-27 |
EP2330505A4 (en) | 2012-08-15 |
EP2330505B1 (en) | 2017-07-05 |
CN102099791A (zh) | 2011-06-15 |
CN102099791B (zh) | 2012-11-07 |
JPWO2010032501A1 (ja) | 2012-02-09 |
JP2011034578A (ja) | 2011-02-17 |
JP4751962B2 (ja) | 2011-08-17 |
US20110113273A1 (en) | 2011-05-12 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP4751962B2 (ja) | 情報処理システムの運用管理方法 | |
Reddy et al. | Metrics for sustainable data centers | |
Ilager et al. | ETAS: Energy and thermal‐aware dynamic virtual machine consolidation in cloud data center with proactive hotspot mitigation | |
Li et al. | Holistic virtual machine scheduling in cloud datacenters towards minimizing total energy | |
US8200995B2 (en) | Information processing system and power-save control method for use in the system | |
Bash et al. | Cool Job Allocation: Measuring the Power Savings of Placing Jobs at Cooling-Efficient Locations in the Data Center. | |
US9389664B2 (en) | Operations management methods and devices thereof in systems | |
Abbasi et al. | Tacoma: Server and workload management in internet data centers considering cooling-computing power trade-off and energy proportionality | |
Tang et al. | Thermal-aware task scheduling for data centers through minimizing heat recirculation | |
US20110131431A1 (en) | Server allocation to workload based on energy profiles | |
Lee et al. | Proactive thermal-aware resource management in virtualized HPC cloud datacenters | |
Varsamopoulos et al. | Energy proportionality and the future: Metrics and directions | |
Li et al. | Data center power minimization with placement optimization of liquid-cooled servers and free air cooling | |
Mukherjee et al. | A detailed study on data centre energy efficiency and efficient cooling techniques | |
JP2011039889A (ja) | ネットワーク運用管理方法及びネットワーク運用管理装置 | |
Khan et al. | Exploratory data analysis for data center energy management | |
Torrens et al. | Integrated Energy Efficient Data Centre Management for Green Cloud Computing-The FP7 GENiC Project Experience | |
Ogura et al. | Model predictive control for energy-efficient operations of data centers with cold aisle containments | |
Moore et al. | Going beyond CPUs: The potential of temperature-aware solutions for the data center | |
EP2575003B1 (en) | Method for determining assignment of loads of data center and information processing system | |
Pesch et al. | Globally optimised energyefficient datacenters | |
JP5165104B2 (ja) | 情報処理システムおよびそのシステムにおける省電力制御方法 | |
Bash et al. | Data center workload placement for energy efficiency | |
JP2013058257A (ja) | 情報処理システムの運用管理方法、運用管理プログラムおよび運用管理装置、ならびに情報処理システム | |
Cao et al. | Green cloud computing with efficient resource allocation approach |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200980127823.8 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09814346 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2010529664 Country of ref document: JP |
|
REEP | Request for entry into the european phase |
Ref document number: 2009814346 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2009814346 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13003687 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |