Method and apparatus for automatically configuring a storage drive array to minimize overheating

Abstract

An information handling device includes a chassis, a motherboard disposed within the chassis, a storage controller, and a plurality of storage drives. The plurality of storage drives are coupled to the storage controller. The storage controller and the storage drives are further coupled to the motherboard. The storage controller is adapted to configure the plurality of storage drives into a storage drive array based on respective physical locations of the plurality of storage drives within the chassis. The devices and methods provided herein allow RAID configuration software to select a set of drives while creating an array with the most diverse cooling zones based on the locations of the drives to minimize the risk of more than one drive in the array failing in quick succession.

Description

Technical Field

[0001] This invention relates to storage devices for information processing equipment, and more particularly to the configuration of storage devices based on their location within the chassis of the information processing equipment. Background Technology

[0002] Servers are widely used by enterprises to provide users with persistent and reliable data processing capabilities, and it is crucial that servers operate 24 / 7 to ensure uninterrupted service at any time. This means that even if a component of the server fails or is damaged, it is desirable that the replacement of that component will not cause any service interruption. Even if there is any service interruption, the affected time should be as short as possible.

[0003] Of all hardware components that fail in servers, disk drives account for nearly 50%, with the remaining half attributable to all other components such as RAM (Random Access Memory), CPU (Central Processing Unit), PSU (Power Supply Unit), NIC (Network Interface Card), fans, system boards, etc. RAID (Redundant Array of Independent Disks) technology is used in almost all production servers to mitigate the common failure of disk drives. However, even with RAID, when one drive fails, it is crucial to quickly swap out and replace the damaged drive, allowing for array rebuilding (which can take hours), while hoping that another drive will not fail during vulnerable periods. This presents challenges in many situations, particularly for hyperscale environments (where too many drives need to be swapped out daily) and IoT (Internet of Things), where servers may be located in remote locations without spare drives. On the other hand, many users in these computing environments may not even use spare drives for RAID configurations in order to save costs and storage space. Summary of the Invention

[0004] Therefore, one aspect of the present invention provides an information processing device, including: a chassis, a motherboard disposed within the chassis, a storage controller, and a plurality of storage drives. The plurality of storage drives are coupled to the storage controller. The storage controller and the storage drives are further coupled to the motherboard. The storage controller is adapted to configure the plurality of storage drives as a storage drive array based on the corresponding physical locations of the plurality of storage drives within the chassis.

[0005] Optionally, the interior of the chassis is divided into multiple cooling zones, and the storage controller is adapted to configure multiple storage drives as a storage drive array based on which cooling zones(s) the multiple storage drives are located in within the chassis.

[0006] Alternatively, each of the multiple cooling zones within the chassis may not overlap. Alternatively, multiple cooling zones may be substantially parallel to each other along the direction from the front to the rear of the chassis.

[0007] Optionally, multiple cooling zones may be defined based on one or more of the following factors: heat dissipation efficiency, estimated operating temperature of the hardware components, and estimated power consumption of the hardware components.

[0008] Optionally, a heat map is generated based on temperature readings from one or more sensors located within the chassis to provide corresponding temperature readings for multiple cooling zones.

[0009] Alternatively, if the storage drives are located in different cooling zones spaced apart from each other, the storage controller is also adapted to configure multiple storage drives as a storage drive array.

[0010] Optionally, if the storage drives are located in a cooling zone with the lowest temperature readings, the storage controller is also adapted to configure multiple storage drives as a storage drive array.

[0011] Optionally, the chassis is adapted to allow multiple storage drives to be connected to at least two of the following: the front drive bay of the chassis, the rear drive bay of the chassis, and the middle drive bay of the chassis.

[0012] Alternatively, the controller is a service processor installed on the motherboard.

[0013] According to another aspect of the present invention, a method for configuring a plurality of storage drives as a storage drive array is provided. The method includes the steps of: a) determining the corresponding physical locations of the storage drives within the chassis of an information processing device; and b) configuring the storage drives as a storage drive array based on the physical locations of the storage devices within the chassis of the information processing device.

[0014] Optionally, step a) further includes: c) dividing the interior of the chassis into multiple cooling zones; and d) determining which cooling zone within the chassis the storage drive is located in.

[0015] Alternatively, each of the multiple cooling zones may not overlap.

[0016] Alternatively, multiple cooling zones may be substantially parallel to each other along the direction from the front to the rear of the chassis.

[0017] Optionally, step d) may also include defining multiple cooling zones based on one or more of the following factors: heat dissipation efficiency, estimated operating temperature of the hardware components, and estimated power consumption of the hardware components.

[0018] Optionally, the method further includes: e) generating a thermal map based on temperature readings from one or more sensors within the chassis to provide corresponding thermal readings for multiple cooling zones.

[0019] Optionally, step b) further includes configuring the storage drives as a storage drive array if the storage drives are located in different cooling zones spaced apart from each other.

[0020] Optionally, step b) further includes configuring multiple storage drives as a storage drive array if the storage drives are located in the cooling zone with the lowest temperature reading.

[0021] Therefore, the embodiments described herein provide a method for RAID configuration software that selects a set of drives while creating an array with the most diverse cooling zones based on drive locations to minimize the risk of more than one drive in the array failing in quick succession. In this way, the information processing device achieves higher storage reliability by automatically assigning storage drives to different disk arrays without any changes to the hardware configuration or the need to purchase additional hard drives for backup. This aims to minimize the risks of concentrating multiple storage drives in a limited number of cooling zones. Users do not need to worry about assigning storage drives to disk arrays or overheating issues, as the software will determine the optimal storage configuration. Instead, users simply need to install the storage drives into the information processing device chassis in the same easy way they would normally. Attached Figure Description

[0022] The foregoing and other features of the present invention will become apparent from the following description of embodiments, which are provided by way of example only in conjunction with the accompanying drawings, wherein: Figure 1 This is a perspective view of a server chassis according to one embodiment.

[0023] Figure 2 It shows the relationship with Figure 1 A block diagram of the server's hardware components related to the control of disk drives in the server.

[0024] Figure 3 This is a flowchart of a method for automatically configuring a storage device for a server according to another embodiment of the present invention.

[0025] Figure 4 This is a 3D heatmap generated for the chassis, illustrated with an example. Figure 3 The method in the middle.

[0026] In the accompanying drawings, similar reference numerals indicate similar components throughout the several embodiments described herein. Detailed Implementation

[0027] Now for reference Figure 1 , Figure 1An information processing apparatus according to an embodiment of the present invention is shown, specifically a server 20. The server 20 includes a chassis 24, which houses most (if not all) of the basic hardware components of the server 20. The server 20 is adapted to process requests and transmit data via a network connection, as understood by those skilled in the art. The server includes a motherboard 25, on which basic components of the server, such as a CPU, logic chipset, expansion cards, network adapters, memory, etc. (all not shown), are hosted. The server also includes a baseboard management controller (BMC) located on the motherboard. Figure 1 (Not shown in the image) is an example of a service processor, which is functionally different from the CPU and can operate independently of the CPU.

[0028] like Figure 1 As shown, the internal space of the chassis 24 is divided into multiple cooling zones 22a-22f. The cooling zones 22a-22f are parallel to each other and do not overlap, and each cooling zone 22a-22f has an elongated shape extending from the front end 24a of the chassis 24 to the rear end 24b of the chassis 24. Therefore, the cooling zones 22a-22f together occupy the entire internal space of the chassis 24. When the cooling zones 22a-22f extend through the entire length of the chassis 24, a front region 26a, a middle region 26b, and a rear region 26c can be further defined within each cooling zone 22a-22f. Therefore, the internal space of the chassis 24 is represented by a 6×3 region matrix, as shown below. Figure 1 As shown.

[0029] Cooling zones 22a-22f are determined based on the hardware design of server 20, and particularly depend on the number of hardware components housed within chassis 24 and their orientation or location. With a fixed hardware configuration of server 20, cooling zones 22a-22f are defined by several factors, including the thermal efficiency of the hardware components within chassis 24, the estimated operating temperature of the hardware components, and the estimated power consumption of the hardware components. In addition to the thermal characteristics of the hardware components, the layout of the airflow paths within chassis 24 is also essential for defining the cooling zones. During the manufacturing of server 20, the vendor of server 20 defines and maintains the layout of cooling zones 22a-22f in the BMC. This pre-configuration is possible because the manufacturing of server 20 means that virtually all hardware components have already been installed in the motherboard 25 and chassis 24. This also takes into account any custom hardware configurations requested by consumers who purchase server 20 from the vendor. Of course, even after server 20 has been sold to and received by a consumer, the consumer can still install or remove certain hardware components, such as expansion cards, within chassis 24. In this case, BMC can continue to use the default cooling zone layout defined for server 20 before leaving the factory, or automatically modify the cooling zone layout based on the identified new installed hardware components and / or any previously removed hardware components, such as information on their constituent elements and the thermal characteristics that may be embedded in the hardware components.

[0030] Thermal maps can be generated for cooling zones 22a-22f (e.g., see [link]). Figure 4 A heat map can be generated using dynamic temperatures at different locations within the chassis, for example, by obtaining readings from sensors (e.g., thermometers) located within the chassis. The heat map is a three-dimensional graph mapping cooling areas to corresponding temperatures within the chassis. The heat map is stored in the BMC as a layout of cooling areas 22a-22f, but it can also be visualized for display to a user on server 20 on a display device (not shown) if needed.

[0031] Chassis 24 allows for the installation of one or more storage drives, although they are not in... Figure 1 As shown in the diagram. For example, at the front end 24a of chassis 24, multiple drive bays 27 are provided to allow user access from the front of chassis 24. Drive bays 27 are adapted to receive storage drives, such as hard disk drives supporting SAS (Serial Attached SCSI) or SATA (Serial Advanced Technology Attached) protocols or SSD drives supporting the Non-Volatile Memory Standard (NVMe) protocol. All drive bays 27 are connected to the drive backplane (…). Figure 1(Not shown in the image), the drive backplane is mounted on the motherboard 25 near the drive bay 27. The drive backplane provides multiple device connectors, allowing one or more storage devices mounted to the drive bay 27 to be connected to the storage controller (…). Figure 1 (Not shown in the diagram). If the drive backplane supports this, hot-swapping of the storage device can be allowed.

[0032] As can be seen, if storage drives are installed into drive bay 27, they will be located in the front range 26a of one or more cooling zones 22a-22f. However, storage drives can also be installed in the middle range 26b and rear range 26c of chassis 24, since middle and rear drive bays (not shown) are also located in chassis 24 next to the front drive bay 27. However, these areas are difficult for users to access, making this less ideal. Storage drives installed in the middle range 26b or rear range 26c can be connected to the same drive backplane as the front storage drives, or they can be connected to different drive backplanes. Users prefer the front range 26a because, for example, when chassis 20 is located in a server rack (not shown), storage drives can be installed therewith without moving chassis 20. In contrast, the middle range 26b requires direct access to the interior of chassis 24 from above. The rear section 26c requires direct access from the rear end 24b of the chassis 24, which means removing the chassis 24 from the server rack and disconnecting all cables and wires connecting the chassis 24 to the rear end 24b.

[0033] Go to Figure 2 The block diagram illustrates various hardware components for electrically connecting to storage drives in server 20 and enabling their data access functionality. A drive backplane 32 is adapted to connect to one or more storage drives 38a-38c. For simplicity, only one drive backplane 32 is shown, but server 20 may contain more than one drive backplane. As those skilled in the art will understand, drive backplane 32 is essentially a component of the storage device connectors, and in this embodiment, there is no bus drive circuitry within drive backplane 32. As a result, data streams from the various device connectors are routed to other cables connecting drive backplane 32 to storage controller 30 in server 20. Depending on the type of switch connector on drive backplane 32, these cables include NVMe cable 28a, SATA / SAS / NVMe X1 cable 28b, and another NVMe cable 28c.

[0034] Driver backplane 32 via I 2The C (internal integrated circuit) bus 35 is connected to the BMC 34. An example of the protocol used by the BMC 34 to communicate with the drive backplane 32 is the Universal Backplane Management (UBM) protocol from the UBM host running on the BMC 34. Therefore, the BMC 34 can read the drive backplane registers on the drive backplane 32 to know how many storage drives are currently installed, their bay numbers, and types (e.g., SATA, SAS, U.2 NVMe drives, or U.3 NVMe drives), collectively referred to as drive information. By storing the layout of the cooling zones in the BMC 34, the BMC 34 then knows which of the cooling zones 22a-22f a particular storage drive 38a, 38b, 38c is located in, and whether the storage drives 38a, 38b, 38c are located in the front range 26a, the middle range 26b, or the rear range 26c. On the BMC 34, RAID configuration software running independently of the CPU or server 20 is installed. The BMC 34 can therefore configure a RAID system using its RAID configuration software by assigning certain storage drives to the RAID array, as those skilled in the art will understand.

[0035] As those skilled in the art will understand, the storage controller 30 is a hardware device or software program for managing a hard disk drive or SSD to operate as a logical unit. In this embodiment, the storage controller 30 is also a RAID controller. The storage controller 30 may be an onboard device on the motherboard, or, more commonly, a PCIe expansion card mounted on the motherboard. The storage controller 30 can be accessed via an I / O controller. 2 The I2C or PCIe bus 36 communicates with the BMC 34. Through the I2C or PCIe bus 36, the BMC 34 can communicate with the storage controller 30 and obtain information about its model, number of PHYs, and connected storage drives. Most importantly, if the user wishes to manually configure RAID functionality, the BMC 34 is suitable for controlling the storage controller 30 to implement a RAID storage drive array, obtaining the status of the RAID array, and providing a user interface for these operations.

[0036] Turn now Figure 3 This demonstrates how the automatic configuration of a RAID disk array on a server works. It should be noted that this configuration method can be applied to... Figure 1-2 The server shown can also be applied to servers with features such as... Figure 1-2 Other information processing devices with different structures are shown. However, for ease of understanding, the following description is based on similar... Figure 1 The layout of the cooling area in the chassis is designed for this purpose.

[0037] The method begins at step 49, where the BMC defines the cooling zones of the server chassis and generates a heat map. As mentioned earlier, cooling zones can be defined and stored in the factory-installed BMC, and can be further updated when the end user makes hardware changes to the server. Once the cooling zones are defined, then... Figure 4 As shown, a heat map 40 is created that associates the six cooling zones within the chassis, and the front, middle, and rear extents are also shown. As described above, the layout of these cooling zones is similar to... Figure 1 As shown, regions 4, 5, and 6, particularly the portions near their rear, have relatively high heat levels, which translates to higher temperatures (represented by peaks in the three-dimensional thermal map), while region 1 represents the lowest heat level among the six cooling regions.

[0038] The method then proceeds to step 50, where it determines the RAID array configuration based on the available storage drives already physically installed on the server. The RAID array can be an existing array containing previously configured storage drives or a new RAID array. As mentioned above, before initiating automatic configuration of the RAID array, the BMC running the RAID configuration software must know all installed storage drives to configure them as a RAID array. The BMC collects information about all installed storage drives, including their type, capacity, cooling zone, and whether they are located in the front, middle, or rear of the chassis. For example, the BMC collects this information when starting the server after a user has installed a new storage drive or removed a previous one. For ease of understanding, let's assume that in this new configuration operation example, the RAID array requires four new drives.

[0039] Then, in step 51, the BMC identifies storage drives of the same type as those already present in the RAID array from all available storage drives. This is because a RAID array requires all storage drives in it to be of the same type (e.g., all SAS, SATA, or U.3 NVMe). Next, in step 52, the capacity of those storage drives identified as available in step 51 is further examined, and only storage drives with a capacity that is the same as or closest to the capacity already present in the RAID array are marked as usable for subsequent steps. Again, this is a well-known requirement for RAID arrays. It should be noted that when the RAID array is newly created, steps 51 and 52 are not required, as there are no existing storage drives in this array that have binding requirements for sequentially added arrays.

[0040] In this example, it is assumed that seven available drives were identified after step 52, and these drives are listed below. These areas refer to one of six cooling zones within the chassis. HL refers to the heat level associated with each drive, and the smaller the number appended to HL, the lower (i.e., cooler) the heat level it indicates (e.g., HL1 means a lower heat level than HL2). Note that the heat level is related to the storage drive, not to the cooling zone, as different parts of a cooling zone may have different heat levels. The front, middle, or rear extent where the storage drive is located, and the cooling zone it occupies, represent the physical location of the storage drive within the chassis.

[0041] Drive 1, HL1 in front of Zone 1 Drive 9, HL2 in front of Zone 3 Drive 10, HL2 in front of Zone 3 Drive 14, HL3 in front of Zone 4 Drive 17, HL4 in front of Zone 5 Drive 22, HL2 in front of Zone 6 Drive 25, HL2 in the middle of Zone 2 The method then continues by selecting four new required drives for the RAID array, one after another. Typically, when selecting drives for a RAID array, the configuration software on the BMC attempts to distribute the drives across as many cooling zones as possible. If the user also configures spare drives, the configuration software selects the spare drives from the least efficient cooling zones, as these are used the least, leaving the spare drives in the more efficient cooling zones for configuration within the RAID array. Thus, this selection is at least in part based on the physical location of the available storage drives relative to the chassis, such as the cooling zones and the front / middle / rear range where the drives are located.

[0042] In step 53, a first drive with the lowest heat level, i.e., drive 1, is selected from the seven available drives. Note that since there is only one available drive with the lowest heat level HL1, no other factors need to be considered when selecting drive 1. However, in other cases, if multiple available drives with the lowest heat levels are available simultaneously, the method will prioritize the drive closest to the front of the chassis, as this means that even if the drive fails during operation, it is easier to replace, minimizing the impact on server operation.

[0043] After selecting drive 1 as the first of four new drives to be configured into the RAID array, the method then proceeds to step 54 to consider whether there are any other new drives to be configured, which is marked "yes" because three more new drives need to be configured. The method then moves to step 55 to select the next drive. The selection of the next drive is essentially similar to that of the first drive, i.e., the drive with the lowest heat level is selected, and the front range is preferred over the middle or rear ranges. However, there is another criterion: in terms of cooling area, it must be far away from the first drive to minimize the chance of both drives failing simultaneously in the event of overheating. In the example described here, after selecting drive 1 with HL1 as the first drive, since there are no more drives with HL1, the next drive is selected from the drives with HL2. Among the drives with HL2, drive 9 in front of area 3 is selected as the second drive, although it should be noted here that drives 10 and 22 also meet the criteria for being selected as the second drive. The method then returns to step 54 to see if more new drives need to be configured. The method then iterates between steps 54 and 55 until all the necessary new drives are configured into the RAID array.

[0044] Drive 22, located in front of Zone 6, is selected as the third drive. As mentioned above, it is in the same position as drive 9, so there is no particular preference for choosing either drive 9 or drive 22 as the second or third drive. Then, for the fourth drive, although drive 10 is located in the front, both drive 10 and drive 25 among the remaining available drives have HL2. Although it might seem that drive 10 should be selected over drive 25 since drive 10 is closer to the front of the chassis, if drive 10 is selected, there is already a drive (drive 9) in the same cooling zone as drive 10. Therefore, to reduce the risk of consecutive failures of storage drives in the same cooling zone, drive 25 in a different cooling zone (Zone 2) is selected as the fourth drive on top of drive 10. At this point, the configuration of all four new drives is complete, and the method proceeds to step 56 to conclude.

[0045] Therefore, exemplary embodiments have been fully described. Although this description relates to embodiments, it will be apparent to those skilled in the art that the invention can be practiced by changing these specific details. Therefore, the invention should not be construed as limited to the embodiments set forth herein.

[0046] Although embodiments have been shown and described in detail in the accompanying drawings and the foregoing description, they should be considered exemplary and not limiting in nature. It should be understood that exemplary embodiments are shown and described only and do not limit the scope of the invention in any way. It will be understood that any feature described herein can be used with any embodiment. The illustrative embodiments are not mutually exclusive, nor do they exclude other embodiments not listed herein. Therefore, the invention also provides embodiments that include combinations of one or more of the illustrative embodiments described above. Modifications and variations can be made to the invention without departing from the spirit and scope thereof; therefore, only the limitations set forth in the appended claims should be applied.

[0047] As an example, in Figure 1 In the embodiments described, the server is used as an example of an information processing device; however, those skilled in the art should understand that in other variations of the invention, different types of information processing devices may be used, such as personal computers, laptops, etc., as long as they contain a service processor and a main processor, wherein the service processor is adapted to automatically configure the operating mode of the storage devices connected to the information processing device.

[0048] exist Figure 1 and Figure 4 In the diagram, the cooling zone is shown as a long, narrow, parallel area within the chassis. However, it should be understood that cooling zones can also take other shapes, depending on the internal space of the chassis and the complexity of the components. For example, one cooling zone could be "T"-shaped, while another could be "L"-shaped. In any case, different cooling zones should not overlap.

[0049] Figure 2 Different cables are shown for connecting the storage controller to the drive backplane, including NVMe cables and SATA / SAS / NVMe x1 cables. However, Figure 2 The illustrations are merely illustrative and there are no limitations on the protocols / types of storage drives used to implement this invention.

Claims

1. An information processing device, comprising: a) The chassis, whose interior is divided into multiple cooling zones; b) The motherboard located inside the chassis; c) Storage controller; as well as d) Multiple storage drives coupled to the storage controller; The storage controller and the storage drive are further coupled to the motherboard; The storage controller is adapted to configure the plurality of storage drives into a storage drive array with the most diverse cooling areas based on which cooling areas(s) ...(s)(s))(s)(s)(s)(s)(s)(s)(s)(s)(s))(s)(s)(s)(s)(s)(s)(s)(s))(s The storage controller is adapted to select a first storage drive with the lowest heat level from the plurality of storage drives for configuration, and after configuring the first storage drive, select a second storage drive with the lowest heat level from the remaining storage drives other than the first storage drive for configuration.

2. The information processing device according to claim 1, further comprising a BMC as a service processor; the BMC being independent of the CPU of the information processing device; the storage controller being adapted to communicate with the BMC; the BMC being adapted to store information related to the plurality of cooling zones and to control the storage controller to configure the plurality of storage drives as the storage drive array.

3. The information processing device according to claim 1, wherein, Each of the plurality of cooling zones within the chassis does not overlap.

4. The information processing device according to claim 1, wherein, The plurality of cooling zones are defined based on one or more of the following factors: heat dissipation efficiency, estimated operating temperature of the hardware components, and estimated power consumption of the hardware components.

5. The information processing device according to claim 1, wherein, A heat map is generated based on temperature readings from one or more sensors located within the chassis to provide corresponding temperature readings for the plurality of cooling zones.

6. The information processing device according to claim 1, wherein, If the storage drives are located in different cooling zones spaced apart from each other, the storage controller is also adapted to configure the plurality of storage drives as a storage drive array.

7. The information processing device according to claim 1, wherein, If the storage drive is located in a cooling zone with the lowest temperature reading, the storage controller is also adapted to configure the plurality of storage drives as the storage drive array.

8. A method for configuring a plurality of storage drives as a storage drive array, comprising the following steps: a) Determining the physical location of the storage drives within the chassis of the information processing equipment, including: i. Divide the interior of the chassis into multiple cooling zones; and ii. Determine which cooling zone(s) within the chassis the storage drive is located in; and its location within those cooling zones. b) Based on the corresponding physical location of the storage drives within the chassis of the information processing device, configuring the storage drives into a storage drive array with the most diverse cooling areas, comprising: i. Select a first storage drive with the lowest heat level from the plurality of storage drives for the configuration; and ii. After configuring the first storage drive, select the second storage drive with the lowest heat level from the remaining storage drives other than the first storage drive among the plurality of storage drives for the configuration.

9. The method according to claim 8, wherein, Step a) is performed by the BMC of the information processing device; the BMC is independent of the CPU of the information processing device; and in step b), the BMC controls the storage controller of the information processing device to configure the storage drive as the storage drive array.

10. The method according to claim 9, wherein, Each of the plurality of cooling zones does not overlap.

Images