Server Thermal Management

Abstract

To provide a server, a method, and a computer readable medium for performing thermal management of a server including a plurality of data storage devices.SOLUTION: A server-based method includes a controller that applies a machine learning model to identify a first portion of a plurality of data storage devices (e.g., SSD) located in a hot zone relative to a second portion of the plurality of data storage devices located outside the hot zone. The controller identifies the hot zone, based on thermal data respectively received from the plurality of data storage devices. Based on the identification of the first portion of the plurality of data storage devices located in the hot zone, the controller performs thermal management of the plurality of data storage devices by logically shuffling the plurality of data storage devices to create a distributed hot zone.SELECTED DRAWING: Figure 6

Description

Technical Field

[0001] (Cross - reference to related applications) This application claims the priority and benefit of U.S. Provisional Patent Application No. 63 / 491,772, filed on March 23, 2023, and the entire content of this provisional patent application is incorporated herein by reference.

[0002] This application generally relates to server thermal management, and more specifically, this application relates to server thermal management using machine learning.

Summary of the Invention

[0003] An enterprise information management system ("IMS") or server can include a chassis system having a plurality of slots for receiving data storage devices such as solid - state drives ("SSDs"). As the density of data storage devices within the IMS and / or server increases, the size of the air channels used to regulate the temperature within the chassis decreases. When multiple data storage devices concentrated in a particular area of the server are operating simultaneously, the data storage devices generate thermal energy, which can form a high - temperature zone within the server. The high - temperature zone within the server can reduce the efficiency of heat transfer from the data storage devices to the external environment and may cause thermal throttling of one or more storage devices.

[0004] Accordingly, in one embodiment, the present disclosure provides a server including a memory configured to store a machine learning model, a plurality of data storage devices disposed within a chassis, and a controller coupled to the memory. The controller receives thermal data related to the plurality of data storage devices and applies the machine learning model to identify a first portion of the plurality of data storage devices located in a high temperature zone relative to a second portion of the plurality of data storage devices located outside the high temperature zone based on the thermal data, and is configured to perform thermal management of the plurality of data storage devices based on the identification of the first portion of the plurality of data storage devices located in the high temperature zone. To perform thermal management, the controller is further configured to logically shuffle the plurality of data storage devices to create a distributed high temperature zone.

[0005] The present disclosure also provides a method for performing thermal management in a server. The method includes receiving thermal data related to a plurality of data storage devices, identifying a first portion of the plurality of data storage devices located in a high temperature zone relative to a second portion of the plurality of data storage devices located outside the high temperature zone based on the thermal data, and performing thermal management of the plurality of data storage devices based on the identification of the first portion of the plurality of data storage devices located in the high temperature zone. Performing thermal management includes logically shuffling the plurality of data storage devices to create a distributed high temperature zone.

[0006] The present disclosure also provides a non-transitory computer-readable medium storing instructions that, when executed by a controller, cause the controller to perform a series of operations, the operations including identifying a first portion of the plurality of data storage devices located in a high temperature zone relative to a second portion of the plurality of data storage devices located outside the high temperature zone based on thermal data, and performing thermal management of the plurality of data storage devices based on the identification of the first portion of the plurality of data storage devices located in the high temperature zone. Thermal management includes logically shuffling the plurality of data storage devices to create a distributed high temperature zone.

[0007] Various aspects of the present disclosure provide improvements in server thermal management. The present disclosure can be embodied in various forms including hardware or circuitry controlled by software, firmware, or combinations thereof. The above summary is intended only to provide a general concept of various aspects of the present disclosure and is not intended to limit the scope of the present disclosure in any way.

Brief Description of the Drawings

[0008]

Figure 1

Figure 2

Figure 3-1

Figure 3-2

Figure 4

Figure 5-1

Figure 5-2

Figure 6

Figure 7-1

Figure 7-2

Figure 8

Figure 9

Figure 10

Figure 11

Figure 12-1

Figure 12-2

Figure 13

Figure 14

Figure 15

Figure 16

Figure 17

DETAILED DESCRIPTION OF THE INVENTION

[0009] In the following description, in order to provide an understanding of one or more aspects of the present disclosure, many details such as data storage device configurations are described. It will be readily apparent to those skilled in the art that these specific details are merely examples and are not intended to limit the scope of the present application. The following description is only intended to provide a general concept of various aspects of the present disclosure and is not intended to limit the scope of the present disclosure in any way. Further, although the present disclosure refers to NAND flash, it will be apparent to those skilled in the art that the concepts discussed herein are applicable to other types of solid-state memories such as NOR, PCM ("phase change memory"), ReRAM, and the like.

[0010] FIG. 1 is a block diagram of a server system 100 according to some embodiments. Although system 100 is described as a server system, it is understood that system 100 may be an IMS system or other system type including multiple data storage devices.

[0011] Server system 100 includes a system memory 102, one or more power supplies 104, an electronic processor 106, a user interface 108, a plurality of data storage devices 110, one or more cooling fans 112 for promoting the flow of air within server system 100, and a communication interface 114. The communication interface 114 can include one or more communication devices such as a network interface device. In one example, various information may be provided to or requested from one or more data storage devices 110 using the communication interface 114. The retrieval of data from and / or the storage of data to the data storage devices 110 may be processed by one or more processors 106. Generally, server system 100 functions as a general server system required for a given application.

[0012] The plurality of data storage devices 110 are disposed within the chassis of server system 100 (e.g., chassis 200 shown in FIG. 2). In one embodiment, the plurality of data storage devices 110 are solid state drives (“SSDs”) such as non-volatile NAND SSDs. However, other SSD types are also contemplated. Further, in other examples, the plurality of data storage devices 110 may be other data storage devices such as hard disk drives (“HDDs”). The cooling fan 112 is configured to direct (e.g., push) air towards one or more of the data storage devices 110. However, in other examples, the cooling fan 112 may be configured to direct air away from (e.g., pull) the data storage devices 110 such that the air flow is pulled across the data storage devices 110 and away from the data storage devices 110.

[0013] Memory 102 stores thermal data 116 associated with the data storage devices 110. For example, memory 102 may periodically receive temperature information regarding the internal temperature of the data storage devices from each of the plurality of data storage devices 110.

[0014] Memory 102 further includes a thermal analysis module 118 for analyzing thermal data 116 and determining corrective measures within server system 100 based on the thermal analysis. In some cases, the thermal data 116 includes the amount of data operations executed by each of the plurality of data storage devices 110, and the electronic processor 106, together with the thermal analysis module 118, performs a thermal analysis of the plurality of data storage devices 110 based on the data operations executed by each of the plurality of data storage devices 110. The electronic processor 106 can determine that a data storage device that executes a large amount of data operations has a higher internal temperature than a data storage device that executes a small amount of data operations.

[0015] The electronic processor 106 identifies the actions of the end user based on the data operation information. Then, the electronic processor 106 determines a control response based on the actions of the end user to maintain efficient heat dissipation in server 100.

[0016] The corrective measures include, for example, logically shuffling the plurality of data storage devices 110 according to the actions of the end user. FIG. 2 shows a first portion of the plurality of data storage devices 110 located in the high temperature zone 304 with respect to a second portion of the plurality of data storage devices located in the low temperature zone 308. FIG. 3 shows a heat map 300a associated with the high temperature zone 304 and the low temperature zone 308 of FIG. 2. The heat map 300a is generated by the electronic processor 106 together with the thermal analysis module 118 and includes respective thermal plots of the plurality of data storage devices 110 according to the respective positions of each data storage device within server 100. For example, a data storage device labeled "#1" is adjacent to a data storage device labeled "#2". Similarly, a data storage device labeled "#2" is also adjacent to a data storage device labeled "#3".

[0017] The example shown in FIGS. 2 and 3 includes 44 data storage devices, but server 100 may include another suitable number of data storage devices. For example, the plurality of data storage devices 110 may include 10 data storage devices, data storage devices, 50 data storage devices, 110 data storage devices, or another suitable number of data storage devices.

[0018] The amount of data operations performed by each of the plurality of data storage devices 110 may vary based on the actions of the end users of each data storage device. Accordingly, heat map 300a provides a comparison of heat data over time for each of the plurality of data storage devices 110.

[0019] In the illustrated example, heat map 300a includes data operation information from 8:00 AM to 5:00 PM, but heat map 300a can include heat data for any suitable period. For example, heat map 300a can include heat data collected over a 24-hour period. The heat data shown in FIG. 3 includes the amount of data transmission measured in terabytes (TB), but alternatively or in addition, may include temperature measurement information from the plurality of data storage devices 110.

[0020] In the example shown in FIGS. 2 and 3, electronic processor 106 can determine that data storage devices #1-#8 are generally more active than data storage devices #9-#44 over a first period (e.g., the hours from 8:00 AM to 5:00 PM). Accordingly, electronic processor 106 can determine that data storage devices #1-#8 define a high temperature zone 304 and data storage devices #9-#44 define a low temperature zone 308. In some cases, electronic processor 106 can determine whether a data storage device is included within high temperature zone 304 by determining whether the data storage device has performed an average amount of data operations greater than an average data operation threshold over a predetermined period.

[0021] In some cases, the electronic processor 106 identifies the high-temperature zone depending on the machine learning model included in the thermal analysis module. The electronic processor 106 can identify the high-temperature zone 304 by determining that at least two adjacent data storage devices exceed a thermal threshold (e.g., an average temperature threshold, an average data operation threshold, etc.). In some cases, the electronic processor 106 identifies the high-temperature zone 304 by determining that another predetermined minimum number of adjacent data storage devices exceeds the thermal threshold. In some cases, the electronic processor 106 identifies the high-temperature zone 304 by determining that at least a minimum number of adjacent data storage devices relative to the total number of data storage devices included in the plurality of data storage devices 110 exceeds the thermal threshold.

[0022] The high-temperature zone 304 can be formed as a result of a set of adjacent data storage devices that perform heat-generating data operations (e.g., a large amount of data reading and data writing) over a period of time. A high-temperature zone containing many high-temperature data storage devices causes inefficient heat dissipation in the server, thereby potentially degrading the performance of the server. Accordingly, in response to the identification of the high-temperature zone 304 within the server 100, the electronic processor 106 performs corrective measures to create a more uniform heat distribution within the server 100. The corrective measures include, for example, logically shuffling the plurality of data storage devices 110 such that the data storage devices performing heat-generating data operations are evenly distributed within the chassis of the server 100. The electronic processor 106 logically shuffles the plurality of data storage devices 110 by exchanging memory locations and data within the plurality of data storage devices 110.

[0023] FIG. 4 shows a first portion of a plurality of data storage devices 110 located in a distributed high temperature zone with respect to a second portion of the plurality of data storage devices 110 located in a distributed low temperature zone. FIG. 5 shows a heat map 300b associated with the distributed high temperature zone and the distributed low temperature zone of FIG. 4. In the example shown in FIGS. 4 and 5, the electronic processor 106 logically shuffles data storage devices #1 - #8 included in the drive positions that define the high temperature zone 304 in the examples of FIGS. 2 and 3, such that data operations performed by these data storage devices are instead performed by data storage devices #1, #8, #13, #19, #26, #32, #39, and #44 at the drive positions, respectively. Similarly, data operations already performed by data storage devices #1, #8, #13, #19, #26, #32, #39, #44 at the drive positions are, after the logical shuffle, performed by data storage devices #1 - #8 at the drive positions, respectively. The electronic processor 106 can perform the logical shuffle such that the data storage device positions that define the high temperature zone are shuffled to a maximum allowable distance from each other (e.g., such that those data storage device positions are evenly distributed across the chassis of the server 100).

[0024] FIG. 6 shows a first exemplary method 600 for performing thermal management in server 100. The electronic processor 106 executes method 600 in conjunction with other components of server 100 (e.g., thermal analysis module 118). FIG. 6 shows steps in a particular order, but in some embodiments, method 600 may be executed in a different order. Further, in some embodiments, method 600 includes additional steps or fewer steps.

[0025] Method 600 includes receiving (at block 604), using the electronic processor 106, thermal data (e.g., thermal data 116) associated with each of the plurality of data storage devices 110. Method 600 includes generating and / or updating (at block 608) a heat map (e.g., heat maps 300a and 300b) based on the thermal data received from each of the plurality of data storage devices 110.

[0026] After updating the heat map, the electronic processor 106 performs a thermal analysis of the heat map (at block 610), using, for example, a machine learning model included in the thermal analysis module 118. Based on the analysis of the heat map, the electronic processor 106 determines (at decision block 612) whether there is a high-temperature zone in the plurality of data storage devices 110.

[0027] If the electronic processor 106 determines (at decision block 612, “YES”) that there is a high-temperature zone in the plurality of data storage devices 110, the electronic processor 106 logically shuffles the plurality of data storage devices 110 to create a dispersed high-temperature zone (at block 614). In contrast, if the electronic processor 106 determines (at decision block 612, “NO”) that there is no high-temperature zone in the plurality of data storage devices, the electronic processor 106 does not logically shuffle the plurality of data storage devices 110 and waits to receive and analyze new SSD thermal data (at block 604).

[0028] Figures 7-9 illustrate a second exemplary use case of the plurality of data storage devices 110 before the electronic processor 106 executes corrective measures. For example, FIG. 7 shows a heat map 700a corresponding to the plurality of data storage devices 110. In the illustrated example, based on the heat map 700a, the electronic processor 106 can determine that a first set 708 of data storage devices including data storage devices #1-#8, #18-#20, #26-#31, and #38-#42 are generally more active during the time from 10:00 AM (UTC-5) to 9:00 PM (UTC-5) than during the time from 9:00 PM (UTC-5) to 10:00 AM (UTC-5). In contrast, the electronic processor 106 can determine that a second set 712 of data storage devices including data storage devices #10-#17, #21-#25, #32-#37, and #43-#44 are generally more active during the time from 9:00 PM (UTC-5) to 10:00 AM (UTC-5) than during the time from 10:00 AM (UTC-5) to 9:00 PM (UTC-5). Accordingly, the electronic processor 106 can determine that the first set 708 of data storage devices execute data operations during the working hours of the end user located in the first time zone, and can determine that the second set 712 of data storage devices execute data operations during the working hours of the user located in a second time zone different from the first time zone.

[0029] Referring now to FIG. 8, the electronic processor 106 can determine that a first set 708 of data storage devices defines a high temperature zone during the hours of 10:00 AM (UTC-5) to 9:00 PM (UTC-5), and a second set 712 of data storage devices defines a low temperature zone during the hours of 10:00 AM (UTC-5) to 9:00 PM (UTC-5). In contrast, as shown in FIG. 9, the electronic processor 106 can determine that a first set 708 of data storage devices defines a low temperature zone during the hours of 9:00 PM (UTC-5) to 10:00 AM (UTC-5), and a second set 712 of data storage devices defines a high temperature zone during the hours of 9:00 PM (UTC-5) to 10:00 AM (UTC-5).

[0030] The electronic processor 106 can logically shuffle a plurality of data storage devices 110 according to the determined pattern. For example, FIGS. 10-12 show a second exemplary use case of a plurality of data storage devices 110 after the electronic processor 106 has executed corrective measures. As shown in FIGS. 10-11 and reflected in the updated heat map 700b of FIG. 12, the electronic processor 106 logically shuffles the plurality of data storage devices 110 such that none of the data storage devices that define a high temperature zone during a predetermined period (e.g., the hours from 10:00 AM (UTC-5) to 9:00 PM (UTC-5)) are directly adjacent to each other.

[0031] FIG. 13 shows a second exemplary method 1300 for logically shuffling a plurality of data storage devices 110. The electronic processor 106 executes the method 1300 together with other components of the server 100 (e.g., the thermal analysis module 118). Although FIG. 13 shows steps in a particular order, in some embodiments, the method 1300 may be executed in a different order. Further, in some embodiments, the method 1300 includes additional steps or fewer steps.

[0032] Method 1300 includes receiving (at block 1304), using the electronic processor 106, thermal data (e.g., thermal data 116) associated with each of the plurality of data storage devices 110. Method 1300 includes generating and / or updating (at block 1306) a heat map (e.g., heat maps 700a and 700b) based on the thermal data received from each of the plurality of data storage devices 110.

[0033] After updating the heat map, the electronic processor 106 performs a thermal analysis of the heat map (at block 1310), for example, using a machine learning model included in the thermal analysis module 118. Method 1300 includes identifying (block 1312) patterns associated with the use of the plurality of data storage devices 110 based on the thermal data 116 associated with the plurality of data storage devices 110. The patterns can include identifying data storage devices that are used more frequently than others during a particular period, identifying data storage devices that are used more frequently during particular days of the week (e.g., weekends and / or weekdays), and / or identifying data storage devices that are used more frequently than others during all periods. The electronic processor 106 can identify a plurality of patterns and logically shuffle the plurality of data storage devices 110 multiple times over a day, week, month, year, or other period.

[0034] After the electronic processor 106 identifies a user pattern, the electronic processor 106 logically shuffles (at block 1314) the plurality of data storage devices 110 according to the user pattern. The electronic processor 106 logically shuffles the plurality of data storage devices 110 such that a set of data storage devices associated with a particular user pattern is such that logical operations performed by the set of data storage devices are not performed by directly adjacent data storage devices.

[0035] Referring now to FIGS. 14 and 15, in some cases, a plurality of data storage devices 110 are disposed within a chassis 1404 of a server 100 that includes one or more air channels 1408. The air channels 1408 can act to improve heat dissipation in the plurality of data storage devices 110, particularly in those data storage devices disposed adjacent to the air channels 1408. As shown in FIG. 14, an electronic processor 106 can identify a high temperature zone 1412 that includes data storage devices #2-#4 and #17-#19. The electronic processor 106 can determine that the air channels 1408 are disposed between data storage devices #6 and #7, between data storage devices #10 and #11, and between data storage devices #14 and #15 (e.g., based on stored information related to the physical configuration of the server 100). Accordingly, the electronic processor 106 can logically shuffle the plurality of data storage devices 110 such that logical operations performed by the data storage devices that define the high temperature zone 1412 are instead performed by data storage devices adjacent to the air channels 1408.

[0036] For example, FIG. 15 shows a plurality of data storage devices 110 having distributed high temperature zones immediately adjacent to the air channels 1408. As shown in FIG. 15, the electronic processor 106 logically shuffles the data storage devices #2-#4 and #17-#19 included in the drive positions that define the high temperature zone 1412 of FIG. 14, such that data operations performed by those data storage devices are instead performed by data storage devices #6 and #7, #10 and #11, and #14 and #15 in the live positions, respectively. Similarly, data operations already performed by data storage devices #6 and #7, #10 and #11, #14 and #15 are respectively performed by data storage devices #2-#4 and #17-#19 in the drive positions after the logical shuffle.

[0037] Figure 16 shows a third exemplary method 1600 for logically shuffling a plurality of data storage devices 110. The electronic processor 106 executes the method 1600 along with other components of the server 100 (e.g., the thermal analysis module 118). Although Figure 16 shows steps in a particular order, in some embodiments, the method 1600 may be executed in a different order. Further, in some embodiments, the method 1600 includes additional steps or fewer steps.

[0038] The method 1600 includes receiving (at block 1604), using the electronic processor 106, thermal data (e.g., thermal data 116) associated with each of the plurality of data storage devices 110. The method 1600 includes generating and / or updating a heat map based on the thermal data received from each of the plurality of data storage devices 110 (at block 1606).

[0039] After updating the heat map, the electronic processor 106 performs a thermal analysis of the heat map using, for example, a machine learning model included in the thermal analysis module 118 (at block 1608). Based on the analysis of the heat map, the electronic processor 106 identifies one or more data storage devices that are used more frequently than the others of the plurality of data storage devices 110 and thus define a high temperature zone (at block 1610).

[0040] After the electronic processor 106 identifies the high temperature zone, the electronic processor 106 logically shuffles the plurality of data storage devices 110 such that data operations performed by the data storage devices included in the high temperature zone are performed by data storage devices disposed directly adjacent to an air channel (e.g., air channel 1408) (at block 1612).

[0041] FIG. 17 shows an exemplary thermal management permission process 1700 executed in server 100. In some cases, electronic processor 106 is configured to provide a notification (e.g., notification 1704) to user interface 108 of server 100 in response to identification of a high temperature zone within a plurality of data storage devices 110 (at block 1708).

[0042] Notification 1704 can include a request generated by electronic processor 106 regarding permission by an operator of server 100 (e.g., operator 1714) to execute corrective measures for thermal management of server 100. In some cases, electronic processor 106 provides notification 1704 to user interface 108 each time a logical shuffle of the plurality of data storage devices 110 is requested. In some cases, electronic processor 106 provides notification 1704 to user interface 108 only once, prior to the first logical shuffle of the plurality of data storage devices 110. In some cases, notification 1704 includes information regarding corrective measures, such as the placement of proposed drive locations. In response to receiving permission from operator 1714 using user interface 108 (at block 1712), electronic processor 106 executes a logical shuffle of a plurality of data storage devices 1716.

[0043] It should be understood that the foregoing description is intended to be illustrative and not restrictive. Many embodiments and applications other than the examples provided will be apparent to those reading the above description. The scope should not be determined with reference to the above description, but instead should be determined with reference to the appended claims, along with the full scope of equivalents to which such claims are entitled. Future developments may occur in the technology discussed herein, and it is expected and intended that the disclosed systems and methods will be incorporated into such future embodiments. In summary, it is understood that the applications are capable of modification and variation.

[0044] All terms used in the claims are intended to be given their broadest reasonable construction and their ordinary meaning as would be understood by one of ordinary skill in the art to which the technology described in this specification pertains, unless an explicit contrary indication is given herein. Specifically, the use of singular articles such as "a", "the", "said", etc. should be read to enumerate one or more of the elements shown, unless an explicit limitation to the contrary is recited in the claim.

[0045] The abstract is provided to enable the reader to quickly ascertain the nature of the technical disclosure. It is understood that the abstract is not used to interpret or limit the scope or meaning of the claims. Additionally, in the [Best Mode for Carrying Out the Invention] described above, it may be seen that various functions are grouped together in various embodiments for the purpose of streamlining the disclosure. The methods of the present disclosure should not be construed as reflecting an intention that the claimed embodiments require more functions than are explicitly recited in each claim. Rather, as reflected in the following claims, the subject matter of the present invention lies in less than all of the functions of the single disclosed embodiment. Accordingly, the following claims are incorporated into the [Best Mode for Carrying Out the Invention], and each claim of the claims stands on its own as a separately claimed subject matter.

Claims

1. A memory configured to store a machine learning model, a plurality of data storage devices disposed within a chassis, and a controller coupled to the memory, the controller analyzes, using a machine learning model, a correlation between a temperature rise and an average data transfer amount from temperature data and average data transfer amounts for each of the plurality of data storage devices over a predetermined period, and identifies a heat-generating data operation that results in a high temperature, thereby predicting a temperature rise for each predetermined period, performs a logical shuffle of memory locations and data within the plurality of data storage devices such that heat generated by the plurality of data storage devices is evenly distributed within the chassis when the predicted temperature rise for each predetermined period exceeds a threshold, Server.

2. Identifying the heat-generating data operation that results in a high temperature includes identifying that at least one of the plurality of data storage devices is used more frequently than other data storage devices during a predetermined period, The server according to claim 1.

3. The chassis includes at least one air channel, and the controller is further configured to exchange the memory location and data of at least one data storage device for which the predicted temperature rise exceeds the threshold with the memory location and data of at least one data storage device disposed directly adjacent to the at least one air channel to perform the logical shuffle of exchanging the memory locations and data within the plurality of data storage devices, The server according to claim 1.

4. further comprising a user interface, the controller provides a notification including a request for permission to perform the logical shuffle to the user interface, and performs the logical shuffle in response to receiving permission from an operator of the server via the user interface, The server according to claim 1, further configured as such.

5. A method for performing thermal management in a server, comprising: analyzing, using a machine learning model, a correlation between a temperature rise and an average data transfer amount from temperature data and average data transfer amounts for each of a plurality of data storage devices disposed within a chassis of the server over a predetermined period, and identifying a heat-generating data operation that results in a high temperature, thereby predicting a temperature rise for each predetermined period, When the predicted temperature rise for each predetermined period exceeds a threshold value, perform a logical shuffle to exchange the memory locations and data within the plurality of data storage devices so that the heat generation by the plurality of data storage devices is evenly distributed within the chassis. A method including the above. **Claim 6**: Identifying the heat-generating data operation that results in high temperature includes identifying that during a predetermined period, at least one of the plurality of data storage devices is used more frequently than other data storage devices. The method according to claim 5. **Claim 7** The chassis includes at least one air channel. The logical shuffle for exchanging the memory locations and data within the plurality of data storage devices includes exchanging the memory locations and data of at least one data storage device whose predicted temperature rise exceeds the threshold value with the memory locations and data of at least one data storage device arranged directly adjacent to the at least one air channel. The method according to claim 5. **Claim 8** Providing a notification including a request for permission to perform a logical shuffle to the user interface of the server. Performing the logical shuffle in response to receiving permission from an operator of the server via the user interface. The method according to claim 5, further including the above. **Claim 9** A non-transitory computer-readable medium storing instructions that, when executed by a controller, cause the controller to perform a series of operations, the operations including: Analyzing the correlation between the temperature rise and the average data transmission volume using a machine learning model based on the temperature data and the average data transmission volume for each predetermined period of a plurality of data storage devices arranged within the chassis of the server, and predicting the temperature rise for each predetermined period by identifying the heat-generating data operation that results in high temperature. When the predicted temperature rise for each predetermined period exceeds a threshold value, perform a logical shuffle to exchange the memory locations and data within the plurality of data storage devices so that the heat generation by the plurality of data storage devices is evenly distributed within the chassis. A non-transitory computer-readable medium including the above. Identifying heat generation data operations that result in high temperatures includes identifying that during a predetermined period, at least one of the plurality of data storage devices is used more frequently than other data storage devices, The non-transitory computer-readable medium according to claim 9.

Images