Site management device and management method

Abstract

To provide a base management apparatus capable of achieving both an improvement in the utilization rate of renewable energy and a reduction in the overhead of power consumption.SOLUTION: A processor 2300 calculates, based on a base power management table 2410, an adjustment amount for adjusting a demand amount consumed at a migration source base from which a computing load is transferred. Based on the adjustment amount and load management information, the processor 2300 selects a migration target load, which is a computing load for transferring an execution base to a migration target base, from among computing loads executed at the migration source base, and generates migration plan information indicating the migration target load.SELECTED DRAWING: Figure 2

Description

Technical Field

[0001] This disclosure relates to a base management device and a management method.

Background Art

[0002] As a countermeasure against climate change, the use of renewable energy is progressing. Conventionally, the procurement method of renewable energy has generally been the purchase of certificates such as non-fossil certificates. However, in recent years, with the increase in the circulation volume of renewable energy, the number of electric power companies that directly purchase renewable energy from power generation companies through PPA (Power Purchase Agreement) and obtain environmental value through the transfer of certificates accompanying the purchase has been increasing. This trend is the same among consumers who consume a large amount of electricity, and in data centers (DCs), which are large consumers of electricity, power procurement through PPA has been spreading mainly in Europe and the United States. In addition, advanced electric power companies in Europe and the United States are considering improving the utilization rate of renewable energy in the whole society in the future, and instead of improving the utilization rate of renewable energy on an annual basis, which is currently mainstream, efforts are being made to improve the utilization rate of renewable energy on a time unit basis.

[0003] On the other hand, since DCs, which are large consumers, require a large amount of power throughout the day, efficient use of renewable energy, which is unstable and has geographical constraints, is an issue. In contrast, a technology that selects a DC that executes a computational load (WL: Workload) according to a region and a time zone rich in renewable energy from a plurality of DCs provided in different locations has attracted attention.

[0004] For example, Patent Document 1 discloses a technique for reallocating an application to any of hosting sites based on the supply amount and cost of power (including power generated by renewable energy) and the available amount of computing resources at each of a plurality of hosting sites.

Prior Art Documents

Patent Documents

[0005] [Patent Document 1] U.S. Patent Application Publication No. 2011 / 0282982 [Overview of the project] [Problems that the invention aims to solve]

[0006] When an application is relocated, the data used by that application must also be copied to the same location as the application. However, this data copying and application switching process results in downtime of several seconds to several minutes, and consequently, power consumption overhead. However, the technology described in Patent Document 1 does not take into account the power consumption associated with application relocation, which may lead to excessive power consumption overhead.

[0007] The object of the present invention is to provide a site management device and management method that can achieve both an increase in the utilization rate of renewable energy and a reduction in power consumption overhead. [Means for solving the problem]

[0008] A site management device according to one aspect of the present disclosure is a site management device that stores data and manages a plurality of sites capable of executing computational loads using said data, the site management device having a processor and a memory, the memory storing power management information that shows for each site a predicted value of the supply amount, which is the amount of electricity supplied by renewable energy to the site, and a predicted value of the demand amount, which is the amount of electricity consumed at the site, the amount of electricity consumed by the computational load, data management information relating to the data used in the computational load, and load management information that shows for each computational load an execution site, which is the site that executes the computational load, the processor calculates an adjustment amount based on the power management information to adjust the demand amount consumed at the first site by moving the execution site from the first site to the second site, and based on the adjustment amount and the load management information, selects a migration target load from among the computational loads executed at the first site that will move the execution site to the second site, and generates migration plan information indicating the migration target load. [Effects of the Invention]

[0009] According to the present invention, it becomes possible to achieve both an increased utilization rate of renewable energy and a reduction in power consumption overhead. This enables a reduction in power consumption and a reduction in applications that degrade service levels in renewable energy-conscious workload control aimed at reducing CO2 emissions. [Brief explanation of the drawing]

[0010] [Figure 1] This figure shows the overall configuration of a multi-site data center management system, which is a computer system according to an embodiment of the present disclosure. [Figure 2] This diagram shows the configuration of the multi-site management server. [Figure 3] This figure shows an example of a site power management table. [Figure 4] This figure shows an example of a placement rules table. [Figure 5] This figure shows an example of an application management table. [Figure 6] This figure shows an example of a storage management table. [Figure 7] This is a flowchart illustrating an example of the migration plan creation process. [Figure 8] This figure shows an example of an interface that displays migration plan information. [Modes for carrying out the invention]

[0011] The embodiments of this disclosure will be described below with reference to the drawings. Note that the embodiments described below are not intended to limit the invention as defined in the claims, and not all of the elements and combinations thereof described in the embodiments are necessarily essential to the solution of the invention.

[0012] Furthermore, in the following explanation, various types of information may be described using the expression "aaa table," but these types of information may also be represented by data structures other than tables. To indicate that the information is not dependent on a data structure, "aaa table" can be called "aaa information." Also, in the following explanation, "program" may be used as the subject, but since a program performs processing defined by execution by a processor (e.g., CPU (Central Processing Unit) or GPU (Graphics Processing Unit)), using memory resources (e.g., memory) and interface devices (e.g., communication devices) as appropriate, the explanation may also be based on the processor. Similarly, the entity performing the processing by executing a program may be a controller, device, system, computer, node, storage device, server, client, or host that has a processor. Also, part or all of a program may be processed using specific hardware circuits. Furthermore, various programs may be installed on each computer by a program distribution server or storage media. Also, in the following explanation, two or more programs may be implemented as a single program, and conversely, one program may be implemented as two or more programs.

[0013] Also, in the following description, although ID is used as the identification information of elements, other types of identification information may be used instead of or in addition to it.

[0014] FIG. 1 is a diagram showing the overall configuration of a multi-site data center management system, which is a computer system according to an embodiment of the present disclosure. The multi-site data center management system shown in FIG. 1 includes a plurality of sites 1000, a multi-site management server 2000, and a client server 3000.

[0015] Each site 1000 is a location where a DC (data center) is arranged. The DC is a group of devices for storing data and executing an application for performing a computational load (computational processing) using the data. Each site 1000 is installed, for example, at geographically separated locations. Also, one site 1000 may have a plurality of DCs. Each site 1000 or each DC may construct different cloud environments such as on-premises, public cloud, and private cloud, respectively. That is, the present disclosure is applicable to a so-called hybrid cloud having a plurality of different cloud environments.

[0016] The multi-site management server 2000 is a site management device (management computer) that manages the execution of applications by each DC and the setting of each device based on information regarding the DCs arranged at each site 1000. The multi-site management server 2000 is communicably connected to the DCs existing at each site 1000 via a wide area network 1800.

[0017] The client server 3000 is a client computer used by a user of the DC and is communicably connected to the DCs existing at each site 1000 via a client network 6000. The client server 3000 transmits various instructions such as an instruction to execute an application to the DC, for example, and receives a response corresponding to the instruction from the DC.

[0018] Each site 1000 has, as components constituting the DC, a DC facility 1050, an IT asset 1100, an application management server 1200, an IT asset management server 1300, and an energy management server 1400, and these are communicably connected to each other via a management network 5000.

[0019] The DC facility 1050 is a facility used for operating the DC, such as, for example, an air conditioning facility and an emergency power supply facility such as a UPS (Uninterruptible Power Supply).

[0020] The IT asset 1100 is a device constituting the DC main body, and is a processing computer for storing data and executing a computational load using the data. In the example of FIG. 1, the IT asset 1100 has a storage 1101 for storing data, a network switch 1102 for managing communication such as data, and a server 1103 for executing an application that performs a computational load using the data.

[0021] The application management server 1200 is a management computer that manages the applications executed by the server 1103 of the IT asset 1100, and acquires and stores application management information regarding the applications.

[0022] The IT asset management server 1300 is a management computer that manages the IT asset 1100, and acquires and stores IT asset information regarding the IT asset 1100. The IT asset information includes, for example, configuration information indicating the connection relationship of each IT asset 1100, and performance information indicate the usage status of each resource (such as a CPU, a network card, and a disk) constituting the IT asset 1100.

[0023] The power management server 1400 is a management computer that manages the power consumption consumed by the DC, and acquires and stores power information related to that power consumption. The power consumption of the DC includes not only the power consumed by IT equipment 1100, but also the power consumed by DC equipment 1050, etc.

[0024] In this embodiment, the management computers (application management server 1200, IT equipment management server 1300, and power management server 1400) are distributed, but any of the management computers may be consolidated with other management computers. Also, in this embodiment, the multi-site management server 2000 is located independently of each site 1000, but it may be located at any of the sites 1000.

[0025] Figure 2 shows the configuration of the multi-site management server 2000. As shown in Figure 2, the multi-site management server 2000 has a Management Network Interface (I / F) 2100, an Input and Output Device (IO) 2200, a Processor 2300, a Local Disk 2400, and Memory 2500.

[0026] The management network interface 2100 is connected to the equipment within each of the 1000 locations shown in Figure 1, enabling communication. The input / output device 2200 is a user interface for inputting and outputting information with the user, such as a monitor, keyboard, and mouse.

[0027] The processor 2300 is a processing unit that performs various operations by loading and executing programs into memory 2500. The local disk 2400 is a storage device that stores programs that define the operation of the processor 2300 and various information used or generated by the processing performed by those programs. The memory 2500 is a main memory that temporarily stores various information and is used as a work area for programs.

[0028] Local disk 2400 stores the following information: Site Energy Management Table 2410, Placement Rule Table 2420, Application Management Table 2430, and Storage Management Table 2440. Local disk 2400 also stores the following programs: Workload Replacement Planning Program 2450 and Workload Management Program 2460.

[0029] The site power management table 2410, the placement rules table 2420, the application management table 2430, and the storage management table 2440 are loaded into memory 2500 and used by the workload migration planning program 2450 and the workload management program 2460.

[0030] The site power management table 2410 is power management information consisting of power demand information for each site 1000 at each time point. The multi-site management server 2000 acquires at least a portion of the power information stored in the power management server 1400 and manages it as the site power management table 2410. The power demand information includes power price information. Furthermore, the power demand information includes not only actual values ​​but also predicted values ​​for power demand. The method of creating the site power management table 2410 is not particularly limited. For example, predicted values ​​may be acquired from the power management server 1400, or they may be calculated from actual values ​​etc. by the multi-site management server 2000.

[0031] The placement rules table 2420, the application management table 2430, and the storage management table 2440 constitute load management information for managing the computing load of the data center.

[0032] The deployment rule table 2420 consists of deployment rule information that indicates deployment rules, which are rules relating to the migration (relocation) of computing load. In this embodiment, the migration of computing load is performed by migrating the application, or more specifically, by migrating the execution site, which is the site 1000 where the application is executed. The method of creating the deployment rule table 2420 is not particularly limited. For example, the deployment rule table 2420 may be created on the multi-site management server 2000, or on another computer such as the IT equipment management server 1300.

[0033] The application management table 2430 consists of application information, which is information about applications executed on the server 1103 of the IT equipment 1100. The multi-site management server 2000 retrieves at least a portion of the application management information stored in the application management server 1200 and manages it as the application management table 2430.

[0034] The storage management table 2440 consists of storage information about storage 1101, which stores data used by the application. The storage information consists of information obtained by the multi-site management server 2000 from the IT equipment management server 1300, but it may also be created from data obtained from other management computers.

[0035] The workload migration planning program 2450 is a program for creating migration plan information that shows the migration plan for the computing load, based on a request from the workload management program 2460. The migration plan shows, for example, the target application, which is the application that runs the target computing load to be migrated, and the destination site 1000 for the target application to be migrated.

[0036] The workload management program 2460 is a program that manages the execution of computing loads. For example, the workload management program 2460 requests the workload migration planning program 2450 to create migration plan information at a predetermined time, and then performs the migration of computing loads based on the migration plan information from the workload migration planning program 2450. The migration of computing loads is performed by migrating the application that performs the computing load. At this time, if the application is using the data area of ​​storage 1101, the data in that data area is migrated (copied) in conjunction with the application during the workload migration.

[0037] Figure 3 shows an example of a site power management table 2410. The site power management table 2410 includes columns 2411 to 2418. The records in the site power management table 2410 show power demand information for a specific site 1000 at a specific time.

[0038] Column 2411 stores the Site ID, which is identification information for location 1000. Column 2412 stores the timestamp (Time). Column 2413 stores the RE Supply Forecast, which is the predicted amount of electricity supplied by renewable energy at the location (location 1000 with the same Site ID in the same record) at the time indicated by the timestamp in the same record. Column 2414 stores the RE Supply Result, which is the actual amount of electricity supplied by renewable energy at the location at the time indicated. Column 2415 stores the RE Price Forecast, which is the price of electricity supplied by renewable energy at the location at the time indicated. Column 2416 stores the GRID Price Forecast, which is the price of electricity supplied by non-renewable energy at the location at the time indicated. Column 2417 stores the Power Demand Forecast, which is the predicted power demand for the site at that time. Column 2418 stores the Power Consumption Result, which is the actual power consumption for the site at that time. Note that the renewable energy supply forecast and power demand forecast represent the supply and demand for that time that were predicted prior to the current time.

[0039] Each record 241A to 241D in the site power management table 2410 shows time-series power demand information for each of the 1000 sites. For example, record 241A shows that at site ID "01" on "2018 / 4 / 1 / 10:00", the predicted renewable energy supply is 3.0 MW, the actual renewable energy supply is 2.5 MW, the renewable energy price is 30 yen / kWh, the non-renewable energy price is 25 yen / kWh, the predicted power demand is 10.1 MW, and the actual power consumption is 10.9 MW. In addition, records such as records 241C and 241D, where the data in columns 2414 and 2418 is "n / a", represent power demand information at a future time.

[0040] Figure 4 shows an example of the placement rule table 2420. The placement rule table 2420 shown in Figure 4 has columns 2421 to 2423.

[0041] Column 2421 stores a Policy ID that identifies the deployment rule. Column 2422 stores the Site IDs of the 1000 locations where the application can be deployed under that deployment rule (the deployment rule for the Policy ID of the same record). Column 2423 stores Migration information indicating whether the application is permitted to migrate between locations 1000 under that deployment rule. In this embodiment, the Migration information shows "True" if migration is permitted and "False" if migration is not permitted.

[0042] Each record 242A to 242E in the deployment rules table 2420 shows the deployment rules information for the application. For example, in record 242A, the deployment rule for policy ID "01" indicates that the application can be deployed to site 1000 with site IDs "01" and "02", and migration between those sites 1000 is permitted. Similarly, in record 242D, the deployment rule for policy ID "04" indicates that the application can be deployed to site 1000 with site IDs "01" and "02", but migration between those sites 1000 is not permitted.

[0043] Figure 5 shows an example of the application management table 2430. The application management table 2430 shown in Figure 5 includes columns 2431 to 2436.

[0044] Column 2431 stores the Application ID, which is identification information for the application. Column 2432 stores the Site ID of the execution site, which is the site 1000 where the application (the application with the same Application ID in the same record) is executed. Column 2433 stores the Estimated Power Consumption, which is an estimated value of the power consumption of the application. Column 2434 stores Storage Configuration information, which shows the configuration of the storage 1101 used by the application. Column 2435 stores the Creation Time, when the application was created (deployed). The creation time may be the time the application was actually created at the site, or, if the application was migrated from another site 1000, the creation time at the previous site 1000 may be carried over. Column 2436 stores the Policy ID of the deployment rules for the application.

[0045] The storage configuration indicates whether a volume is allocated to the application (i.e., whether data used by the application exists) and whether the data has been copied. Specifically, the storage configuration information shows "No Volume" if no volume is allocated to the application, and "Allocated" if a volume is allocated to the application. Furthermore, the storage configuration information shows "Allocated and Copied for Migration" if a volume is allocated to the application and data has been copied for application migration, and "Allocated and Copied for DR" if a volume is allocated to the application and data has been copied for application disaster recovery. Note that data copying may also be performed for other purposes such as backup.

[0046] Each record 243A to 243D in the application management table 2430 shows application information. For example, record 243A indicates that application ID "1" is running at site ID "01", has an estimated power consumption of "30wh", has no volume allocated for storage, has been running since "2018 / 4 / 1 / 10:00", and follows the placement rules of policy ID "01".

[0047] Figure 6 shows an example of the storage management table 2440. The storage management table 2440 shown in Figure 6 includes columns 2441 to 2449.

[0048] Column 2441 stores the Volume ID, which is identification information for identifying a volume. Column 2442 stores the Application ID of the application to which the volume (the volume with the Volume ID included in the same record) is assigned. Column 2443 stores the Copy Flag, which indicates whether the volume is in a copy state, meaning its data is being copied to another volume. In this embodiment, the Copy Flag is "True" if the volume is in a copy state, and "False" if it is not. Column 2444 stores the Copy Pair ID, which is identification information for identifying the copy pair of the volume. A copy pair is a combination of a source volume and a destination volume in a copy state, and if the volume is not in a copy state, the Copy Pair ID is "n / a". Column 2445 stores the Type of the volume copy pair. The Type is "Primary" if the volume is the source volume, and "Secondary" if the volume is the destination volume. Column 2446 stores the Capacity of the volume. Column 2447 stores the used capacity of the volume. The sum of the used capacities of the volumes allocated to the application represents the amount of data used by the application. Column 2448 stores the device type, which is the type of device (storage 1101) to which the volume is allocated. Column 2449 stores the site ID of the location 1000 that has the volume.

[0049] Records 244A to 244E in the storage management table 2440 show storage information. For example, record 244A shows that volume ID "1" is assigned to application ID "02", that the volume is in a copy state, has copy pair ID "01", is a primary volume, has a capacity of "500GB" of which "110GB" is used, has a device type of "A", and is located at site 1000 with site ID "01". Note that the volumes corresponding to records 244A and 244B each form a copy pair with each other. Record 244E shows a volume that does not form a copy pair.

[0050] Furthermore, the storage configuration information stored in column 2434 of the application management table 2430 and the storage management table 2440 constitute data management information related to the data used by the application.

[0051] The following describes in more detail the migration plan creation process using the workload migration planning program 2450. The migration plan creation process is the process of creating migration plan information based on the site power management table 2410, the placement rules table 2420, the application management table 2430, and the storage management table 2440.

[0052] Figure 7 is a flowchart illustrating an example of the migration plan creation process.

[0053] In step S1, the workload migration planning program 2450 receives a request from the workload management program 2460 to create migration plan information and starts the following process.

[0054] In step S2, the workload migration planning program 2450 obtains the predicted renewable energy supply and power demand values ​​for each site 1000 at each time from the site power management table 2410.

[0055] In step S3, the workload migration planning program 2450 calculates an adjustment amount to adjust the amount of electricity consumed at each site 1000 at predetermined intervals within a predetermined period, based on the renewable energy supply forecast and electricity demand forecast obtained in step S2 and the placement rule table 2420. The adjustment of electricity demand is performed by reallocating the computational load, as described later. In this embodiment, the predetermined time is one hour, which is the time interval of the timestamps in the site power management table 2410 shown in Figure 4, but it is not limited to this example. The period is, for example, one day.

[0056] In calculating the adjustment amount, if the connection relationship between the 1000 sites is one-to-one, the workload transition planning program 2450 first calculates the surplus of renewable energy supply, which is the amount of electricity supplied by renewable energy, by subtracting the predicted electricity demand value PDFt from the predicted renewable energy supply value RSFt at each 1000 site (RSFt-PDFt). Hereafter, the surplus of renewable energy supply will be referred to as the renewable energy surplus. Also, if the renewable energy surplus is negative, the absolute value of the renewable energy surplus may be referred to as the deficit.

[0057] If the renewable energy surplus at one site 1000 is positive and the renewable energy surplus at the other site 1000 is negative, the workload migration planning program 2450 determines that there is a shortage of renewable energy supply at the other site 1000 and sets one site 1000 as the source site (first site) and the other site 1000 as the destination site (second site). Then, if the renewable energy surplus at the source site is greater than the shortage of renewable energy supply at the destination site, the workload migration planning program 2450 uses the shortage as the adjustment amount. If the renewable energy surplus at the source site is less than the shortage of renewable energy supply at the destination site, the program uses the renewable energy surplus as the adjustment amount.

[0058] Similarly, if the connection relationship between the 1000 sites is n to n, the workload migration planning program 2450 designates the 1000 sites with a negative renewable energy surplus as source sites and the 1000 sites with a positive renewable energy surplus as destination sites. In this case, the workload migration planning program 2450 may set the adjustment amount for each destination site to a value obtained by dividing the deficit of the source sites equally among the number of destination sites, or to a value corresponding to the renewable energy surplus of the destination sites.

[0059] In step S4, the workload migration planning program 2450 starts a loop process that repeats the following steps S5 to S14 for each combination of sites that adjust the power of the adjustment amount calculated in step S3.

[0060] In step S5, the workload migration planning program 2450 sorts the records in the application management table 2430 for selected applications that run on the source site (the site with a negative renewable energy surplus) among the given site combinations, in descending order of the estimated power consumption of the selected applications.

[0061] In step S6, the workload migration planning program 2450 refers to the placement rules table 2420, the application management table 2430, and the storage management table 2440 to select from the list of target applications applications that are migratable and have no data allocated to them, and applications that are migratable and have already had their data copied to the destination site, as candidate applications for migration. Whether an application is migratable or not is determined based on the migration feasibility information stored in column 2423 of the placement rules table 2420. Whether data has been allocated or not is determined based on the storage configuration information stored in column 2434 of the application management table 2430. Specifically, if the storage configuration information for an application indicates "No Volume," it is determined that no data has been allocated; otherwise, it is determined that data has been allocated. Whether data has been copied or not is determined based on whether the destination site, which is the location of the location ID of the volume with type "Secondary" stored in column 2445 of the storage management table 2440, is the destination site. If the destination site is the destination site, it means that the copied data has already been stored at the destination site.

[0062] In step S7, the workload migration planning program 2450 refers to the application management table 2430 to calculate the total power consumption of the candidate applications selected in step S6, and determines whether this total value is greater than or equal to the adjustment amount calculated in step S3. If the total value is greater than or equal to the adjustment amount (step S7: Yes), the workload migration planning program 2450 proceeds to step S11; if the total value is less than the adjustment amount (step S7: No), it proceeds to step S8.

[0063] In step S8, the workload migration planning program 2450 refers to the application management table 2430 to identify applications from the selected applications that are migrateable, have data allocated to them, and whose data has not yet been copied to the destination site. Based on the storage management table 2440, the workload migration planning program 2450 calculates the copy cost, which is the cost of copying the data allocated to the identified applications to the destination site. The copy cost is a value calculated based on the used capacity, i.e., the amount of data, stored in column 2447 of the storage management table 2440, and is, for example, at least one of the power consumption and time involved in the copy. For example, the workload migration planning program 2450 refers to the device type at the destination site stored in column 2448 of the storage management table 2440 and calculates the copy cost from the amount of data using a predetermined formula for that device type. However, the method of calculating the cost is not limited to this method.

[0064] In step S9, the workload migration planning program 2450 selects as candidate applications from the applications identified in step S8 that meet predetermined criteria for copy cost. The criteria are, for example, that the cost is below a threshold. The threshold may be a fixed value or a variable value. A variable threshold may be a percentage of a variable criterion. For example, the threshold for copy power consumption may be 5% of the adjustment amount, and the threshold for copy time may be 10% of the time until the scheduled migration time. The workload migration planning program 2450 may also select applications where at least one of copy power consumption and copy time is below the threshold, or where both copy power consumption and copy time are below the threshold.

[0065] In step S10, the workload migration planning program 2450 refers to the application management table 2430 and selects from the applications identified in step S8, among the candidate applications, applications whose lifetime (the period during which an application is continuously running) is longer than the reference period, as candidate applications. The reference period may be a fixed value, such as 24 hours, or a variable value.

[0066] In step S11, the workload migration planning program 2450 identifies candidate applications to be migrated, starting with the top records sorted in step S5 (i.e., applications with the highest power consumption), until the total power consumption of those applications reaches the adjustment amount determined in step S3.

[0067] In step S12, the workload migration planning program 2450 predicts, based on the site power management table 2410, whether the application to be migrated may be moved back (relocated) from the destination site (second site) to the source site (first site) during a predetermined period after the scheduled migration time. For example, the workload migration planning program 2450 calculates the adjustment amount between sites 1000 as performed in step S3 for the time within the predetermined period, and based on that adjustment amount, determines whether the application will be migrated from the second site to the first site in the same combination of sites as the one in question. For example, if it is determined that 100 kW of application will be migrated from the first site to the second site after 1 hour, and then another 500 kW of application will be migrated from the second site to the first site after 10 hours, the workload migration planning program 2450 predicts that the application may be moved back.

[0068] The workload migration planning program 2450 proceeds to step S13 if it is expected to return (step S12: Yes), and to step S14 if it is not expected to return (step S12: No).

[0069] In step S13, the workload migration plan program 2450 develops a deletion policy that does not delete volumes at the source site when copying data during application migration.

[0070] In step S14, the workload migration plan program 2450 develops a deletion policy to delete volumes at the source site when data is copied during application migration.

[0071] Then, after the workload migration planning program 2450 has performed steps S5 to S14 for all 1000 sites, it terminates the loop process in step S15.

[0072] Furthermore, in step S16, the workload migration planning program 2450 returns migration plan information, including the application to be migrated, its destination location, and its deletion policy, to the workload management program 2460, and then terminates the process.

[0073] In the above process, the workload migration planning program 2450 would perform the subsequent processing after receiving instructions from the workload management program 2460 in step S1. However, the workload migration planning program 2450 may pre-generate a list of applications to be migrated by executing the processes in steps S2 to S11 in advance. In this case, the workload migration planning program 2450 may return migration plan information to the workload management program 2460 based on that list after receiving instructions from the workload management program 2460.

[0074] Furthermore, while the above process only selects sites with a positive renewable energy surplus as destination sites, if a site has a negative renewable energy surplus but prioritizes economic benefits by sharing renewable energy power with other sites, that site with the negative renewable energy surplus may also be selected as a destination site. In this case, for example, information indicating whether or not renewable energy power will be shared for each site is added to the site power management table 2410, and the workload migration planning program 2450 selects a destination site based on that information.

[0075] Furthermore, not all of the above processes are necessarily required, and some can be omitted as appropriate. For example, the process in step S10 may be omitted.

[0076] Furthermore, when the workload management program 2460 receives migration plan information from the workload migration planning program 2450, it presents that migration plan information to the user. Then, when the user instructs the workload management program 2460 to migrate the application according to the migration plan information, it instructs each site 1000 to carry out the migration of the target application according to that migration plan information. At that time, the workload management program 2460 sends an instruction to delete the volume to the source site of the target application whose migration plan information specifies a deletion policy of deleting volumes, and deletes the volume after the data has been copied. On the other hand, for the source site of the target application whose deletion policy does not specify a deletion policy of not deleting volumes, the workload management program 2460 does not send an instruction to delete the volume, and leaves the volume after the data has been copied. Note that when a volume is deleted, the data assigned to that volume is also deleted.

[0077] Figure 8 shows an example of a GUI (Graphical User Interface) that the workload management program 2460 uses to present migration plan information to users. The GUI 9000 shown in Figure 8 is presented, for example, to a DC user.

[0078] GUI9000 has display units 9100, 9200, and 9300. Display unit 9100 shows a list of the configurations of DCs and base stations 1000 managed by the user. In the example in Figure 8, display unit 910 shows that there are base stations "01" and "02", and that base station "01" includes data centers "DC1", "DC2", and "DC3".

[0079] The display unit 9200 displays a chart 9210 showing the predicted renewable energy supply and electricity demand values ​​for the target period (one day in the diagram) at the location selected by the display unit 9100. Chart 9210 shows the predicted electricity demand with a dashed line and the predicted renewable energy supply with a solid line. In addition, the display unit 9100 displays shortage information 9211 to the left of chart 9210, indicating the time periods and amounts of shortages where the predicted renewable energy supply is insufficient compared to the predicted electricity demand.

[0080] The display unit 9300 displays transition plan information to address the shortage of electricity from renewable energy sources.

[0081] As described above, according to this embodiment, the site power management table 2410 shows, for each site 1000, a predicted renewable energy supply value, which is the predicted amount of electricity supplied by renewable energy to each site 1000, and a predicted electricity demand value, which is the predicted amount of electricity consumed at site 1000. The load management information (placement rule table 2420, application management table 2430, and storage management table 2440) shows, for each computing load, the amount of electricity consumed by the computing load, data management information regarding the data used by the computing load, and the execution site, which is site 1000 that executes the computing load. The workload migration planning program 2450 calculates an adjustment amount to adjust the demand consumed at the source site of the computing load migration, based on the site power management table 2410. Based on the adjustment amount and load management information, the workload migration planning program 2450 selects the target loads for migration from among the computing loads executed at the source site, which are computing loads whose execution site will be moved to the destination site, and generates migration plan information indicating the target loads for migration. Therefore, it becomes possible to generate migration plan information that takes into account information about the data used by the application, thereby minimizing various problems associated with site relocation, such as power consumption overhead. It becomes possible to achieve both increased utilization of renewable energy and reduced power consumption overhead. Consequently, it becomes possible to reduce power consumption and applications that experience a decline in service levels in renewable energy-conscious workload control aimed at reducing CO2 emissions.

[0082] Furthermore, in this embodiment, the workload migration planning program 2450 selects candidate loads to be migrated based on load management information, and selects candidate loads as migrated loads in descending order of power consumption until the total power consumption of the candidate loads reaches the adjustment amount. In this case, it is possible to select migrated loads in a way that minimizes the number of computing loads to be migrated, thereby reducing the number of computing loads (applications) that are stopped during the migration, and thus enabling a more appropriate balance between improving the utilization rate of renewable energy and reducing power consumption overhead.

[0083] Furthermore, in this embodiment, the workload migration planning program 2450 selects as candidate loads computing loads where the destination site for data copying is the destination site for migration. Therefore, it becomes unnecessary to copy data for the migration of computing loads, thus minimizing the problem of data migration (copying) failure.

[0084] Furthermore, in this embodiment, the workload migration planning program 2450 selects computational loads for which no data is available as candidate loads. Therefore, it is not necessary to copy data for the migration of computational loads, which further reduces power consumption overhead.

[0085] Furthermore, in this embodiment, the workload migration planning program 2450 selects as candidate loads computational loads for which the cost (e.g., power consumption and time) of copying data to the destination site is below a threshold. Therefore, since computational loads with low migration costs can be selected as candidate loads, it becomes possible to further reduce power consumption overhead.

[0086] Furthermore, in this embodiment, the workload migration planning program 2450 selects as candidate workloads that have a lifetime of continuous execution equal to or greater than the reference period. In this case, it is possible to suppress the wasteful consumption of migration costs by eliminating workloads with short lifetimes immediately after migration.

[0087] The Workload Migration Planning Program 2450 retains data at the destination site without deleting it if it is expected that the computing load will return from the destination site to the source site. This eliminates or reduces the need to copy data when the computing load returns, thereby further reducing power consumption overhead.

[0088] The embodiments of the Disclosure described above are illustrative for illustrative purposes and are not intended to limit the scope of the Disclosure to those embodiments only. Those skilled in the art can implement the Disclosure in various other forms without departing from the scope of the Disclosure. [Explanation of Symbols]

[0089] 1000: Site 1050: DC Equipment 1100: IT Equipment 1101: Storage 1103: Server 1200: Application Management Server 1300: IT Equipment Management Server 1400: Power Management Server 1800: Wide Area Network 2000: Multi-Site Management Server 2100: Management Network I / F 2200: Input / Output Device 2300: Processor 2400: Local Disk 2410: Site Power Management Table 2420: Placement Rules Table 2430: Application Management Table 2440: Storage Management Table 2450: Workload Migration Planning Program 2460: Workload Management Program 2500: Memory 3000: Client Server

Claims

1. A site management device that stores data and manages multiple sites capable of performing computational loads using said data, The aforementioned site management device has a processor and memory, The aforementioned memory is Power management information that shows, for each of the aforementioned locations, a predicted value of the supply amount, which is the amount of electricity supplied by renewable energy to the aforementioned location, and a predicted value of the demand amount, which is the amount of electricity consumed at the aforementioned location. The system stores the amount of power consumed by the computing load, data management information relating to the data used by the computing load, and load management information indicating the execution site where the computing load is executed for each computing load. The aforementioned data management information indicates the copy destination location, which is the location where the copy data of the aforementioned data is stored, and whether or not the aforementioned data exists. The aforementioned processor, Based on the aforementioned power management information, an adjustment amount is calculated to adjust the amount of demand consumed at the first location by shifting the execution location from the first location to the second location. Based on the load management information, at least one of the computing loads, specifically the computing load without data and the computing load whose copy destination is the second location, is selected as a candidate load to be migrated, which is a computing load whose execution location is migrated to the second location. A site management device that selects a load to be migrated from among the candidate loads based on the adjustment amount and the amount of power consumed by the candidate load, and generates migration plan information indicating the load to be migrated.

2. The site management device according to claim 1, wherein the processor selects the candidate loads as the loads to be migrated in descending order of their power consumption until the total power consumption of the candidate loads reaches the adjustment amount.

3. The aforementioned data management information indicates the amount of data, The site management device according to claim 1, wherein the processor calculates the cost of copying the data from the first site to the second site based on the amount of data, and selects the computation load for which the cost is below a threshold as the candidate load.

4. The site management device according to claim 3, wherein the aforementioned cost is at least one of the amount of electricity and time.

5. The load management information further indicates the lifetime, which is the period during which the computational load is continuously executed. The site management device according to claim 1, wherein the processor selects the computational load whose lifespan is equal to or greater than the reference period as the candidate load.

6. The aforementioned processor, The transition plan information for a predetermined time is generated, Based on the power management information, it is predicted whether the execution site will be returned from the second site to the first site during a predetermined period after the predetermined time. If it is anticipated that the aforementioned execution site may be returned, based on the aforementioned transition plan information When the execution site is migrated, the data stored at the first site will be retained. If it is not expected that the execution site will be returned, the data stored at the first site will be deleted when the execution site is migrated based on the migration plan information. The location management device according to claim 1.

7. A management method using a site management device that stores data and manages multiple sites capable of performing computational loads using said data. The aforementioned site management device has a processor and memory, The aforementioned memory is Power management information that shows, for each of the aforementioned locations, a predicted value of the supply amount, which is the amount of electricity supplied by renewable energy to the aforementioned location, and a predicted value of the demand amount, which is the amount of electricity consumed at the aforementioned location. The system stores the amount of power consumed by the computing load, data management information relating to the data used by the computing load, and load management information indicating the execution site where the computing load is executed for each computing load. The aforementioned data management information indicates the copy destination location, which is the location where the copy data of the aforementioned data is stored, and whether or not the aforementioned data exists. The aforementioned processor, Based on the aforementioned power management information, an adjustment amount is calculated to adjust the amount of demand consumed at the first location by shifting the execution location from the first location to the second location. Based on the load management information, at least one of the computing loads, specifically the computing load without data and the computing load whose copy destination is the second location, is selected as a candidate load to be migrated, which is a computing load whose execution location is migrated to the second location. A management method that selects a load to be migrated from among the candidate loads based on the adjustment amount and the amount of power consumed by the candidate load, and generates migration plan information indicating the load to be migrated.

Images