WO2023035113A1

WO2023035113A1 - Control method and apparatus for storage apparatus

Info

Publication number: WO2023035113A1
Application number: PCT/CN2021/116983
Authority: WO
Inventors: 底晔佳; 王金伟; 黄恩走
Original assignee: 华为技术有限公司
Priority date: 2021-09-07
Filing date: 2021-09-07
Publication date: 2023-03-16
Also published as: CN117769699A

Abstract

Disclosed in the present application are a control method and apparatus for a storage apparatus. The storage apparatus comprises a controller and a storage array. The method is applied to the controller. The method comprises: when it is predicted that there is burst traffic within a first time period, increasing, before the first time period, bandwidth resources that are allocated for a garbage collection (GC) operation, wherein the burst traffic includes a first number of data processing requests, the first number being greater than the number of data processing requests to which a storage apparatus can respond per second; and executing the GC operation by using a first bandwidth resource, so as to obtain a first available storage space in a storage array, wherein the first bandwidth resource is a bandwidth resource that is allocated for the GC operation. In the solution, the execution of a GC operation is promoted before the arrival of burst traffic, such that a relatively large available storage space can be obtained, and when the burst traffic arrives, it can be ensured that there is a relatively large available storage space in a storage apparatus for responding to the burst traffic, thereby improving the stability and input/output (IO) response speed of the storage apparatus.

Description

Method and device for controlling a storage device

technical field

The present application relates to the field of storage technology, and in particular, to a method and device for controlling a storage device.

Background technique

Solid state disk (SSD) is now widely used in various storage scenarios, such as servers, data centers, embedded devices, and so on.

SSD is equipped with garbage collection GC (garbage collection, GC) algorithm. The GC algorithm is used to perform garbage collection on the storage space of stored data when the available space (that is, empty space) in the SSD storage space is less than a threshold (called GC waterline), thereby releasing some storage space.

After a user request (IO) is sent to the SSD, the SSD needs a certain amount of time to respond to the IO. When the response time exceeds the user tolerance threshold (such as 1s), the user experience will be affected. Wherein, the response time of the IO is the total time from when the IO is issued to when the IO is processed, and the response time includes not only the time for flash memory processing, but also the time for waiting for the IO to be processed. When the SSD continues to respond to the received IO stably, the number of IOs processed by the SSD per second (IOPS) is called the disk capacity. Generally, due to the resources occupied by GC, the disk capacity of SSD will be less than its maximum capacity.

Currently, in SSD control, the resources used by GC vary with the resources used by IO responses. When more resources are used by IO responses, more resources are allocated to GC. When fewer resources are used by IO responses, There are also fewer resources allocated for GC use. Among them, when the amount of IO access is small, the resources and storage space required by SSD to respond to IO, and the storage space occupied by the resulting garbage are all small, and the storage space occupied by IO response can be maintained by using less resources for GC The relative stability of storage space released by GC. Therefore, when the IO access volume is less than the capacity of the disk (called regular traffic), the SSD can complete the IO on time and at the same time perform GC normally to maintain a stable available storage space. However, when the IO access volume exceeds the capacity of the disk (called burst traffic), the resources required for processing IO and the consumed storage space will greatly increase in a short period of time. On the one hand, the GC may not be able to reclaim enough storage space in a short period of time, which will break the relative stability between the storage space occupied by the IO response and the storage space released by the GC, and may even cause the storage space to be consumed by the IO response process, affecting Stability of the storage system. On the other hand, GC also needs to use more resources to speed up the release of storage space, so as to release as much storage space as possible for IO response, which will lead to resource contention between GC and IO response process, affecting the IO response speed, which may lead to IO response timeout problems.

Contents of the invention

The present application provides a storage device control method and device, which are used to improve the stability and IO response speed of the storage device by controlling resource allocation and execution of garbage collection GC operations in the storage device.

In a first aspect, the present application provides a method for controlling a storage device, the storage device includes a controller and a storage array, the method is applied to the controller, and the method includes: when it is predicted that the memory During burst traffic, before the first time period, increase the bandwidth resources allocated to garbage collection GC operations; wherein, the burst traffic includes a first number of data processing requests, and the first number is greater than the The number of data processing requests that the storage device can respond to per second; use the first bandwidth resource to execute the GC operation to obtain the first available storage space in the storage array, wherein the first bandwidth resource is allocated to the Bandwidth resources for GC operations.

In this method, when the controller of the storage device predicts the burst traffic, by increasing the bandwidth resources allocated to the GC operation before the burst traffic arrives, the execution speed of the GC operation can be accelerated, so that the GC operation can be performed in a short time Fast reclamation within the system will result in more available storage space. Therefore, this method can be used to prepare storage space in advance for responding to burst traffic when burst traffic is predicted. In this way, when the burst traffic comes, it can ensure that there is more storage space in the storage device to respond to the burst traffic, thereby ensuring a high response speed while avoiding excessive consumption of available storage space and improving the stability of the storage device .

In a possible design, the method further includes: when it is determined that the burst traffic exists, allocating the first available storage space to an IO operation corresponding to the burst traffic.

In this method, when it is determined that there is a burst traffic, the first available storage space obtained through advance processing of the GC operation is allocated to the IO operation corresponding to the burst traffic, which can ensure more storage space for the IO operation. On the one hand, because the storage space is relatively sufficient, the execution speed of the IO operation can be improved; on the other hand, the excessive consumption of the available storage space can be avoided, and the stability of the storage device can be improved.

In a possible design, when it is predicted that there is a burst traffic within the first time period, the method further includes: increasing the start threshold of the GC operation; wherein, when the available storage in the storage array When the space is smaller than the start threshold, start the GC operation.

In this method, by increasing the startup threshold of the GC operation when the burst traffic is predicted, the controller can obtain a larger available storage space by executing the GC operation, and it can also ensure that the controller obtains a large storage space by executing the GC operation. After a larger available storage space, the available storage space can be maintained by performing GC operations.

In a possible design, when it is predicted that there is a burst traffic within the first time period, the method further includes: based on the storage space required for the IO operation corresponding to the burst traffic and the first bandwidth Resources determine a first target time; when the first target time arrives, increase the startup threshold of the GC operation; wherein, when the available storage space in the storage array is less than the startup threshold, start the GC operate.

In this method, according to the storage space required for the IO operation corresponding to the burst traffic and the bandwidth resources allocated to the GC operation, the time required for the GC operation to process and obtain the storage space can be determined. Further, this time can be used as a reference to further Accurately determining the time to increase the start threshold of the GC operation can avoid raising the start threshold of the GC operation too early or too late, thereby reducing unnecessary GC operations and reducing the impact on the life of the storage device, and can be obtained through the GC operation. Large enough available storage space for burst traffic.

In a possible design, the method further includes: when it is determined that the burst traffic exists, lowering the start threshold of the GC operation; or, if the predicted burst traffic does not exist at the second target moment When the GC operation arrives, lower the start threshold of the GC operation; wherein, the second target time is any time within the first time period.

In this method, when it is determined that the burst traffic exists, the GC operation can be stopped by lowering the start threshold of the GC operation, so that more bandwidth resources can be released for use by the IO operation, and the IO operation can be timely and quickly Respond to burst traffic, thereby reducing or avoiding the problem of IO response timeout. When the predicted burst traffic does not exist, if a high GC operation start threshold is maintained all the time, the life span of the storage device will be affected due to excessive and unnecessary GC operations. Therefore, reducing the start threshold of the GC operation when it is determined that the predicted burst traffic does not come, can reduce the impact on the life of the storage device.

In a possible design, the method further includes: before it is determined that the burst traffic exists, when the first available storage space reaches a target value, allocating the second bandwidth resource to the storage space corresponding to the current data processing request IO operation, and assigning a third bandwidth resource to the GC operation; wherein, the target value is greater than or equal to the storage space required for the IO operation corresponding to the burst traffic; the size of the second bandwidth resource is the same as the The size of the available storage space in the storage array has a positive correlation, and the size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.

In this method, when the available storage space in the storage array reaches the target value, it indicates that the first available storage space obtained through the GC operation is sufficient for the IO operation corresponding to the burst traffic. If it is not confirmed that there is a burst traffic at this time, by allocating bandwidth resources that have a positive correlation with the available storage space for IO operations, and allocating bandwidth resources that have a negative correlation with the available storage space for GC operations, you can allocate bandwidth resources in the available storage space When it is too large, it will speed up the consumption of storage space and slow down the recovery of available storage space. When the available storage space is too small, it will slow down the consumption of storage space and speed up the recovery of storage space, so that the size of available storage space remains relatively Stable to wait for the arrival of burst traffic.

In a possible design, the second bandwidth resource complies with the following formula:

BW ₁ =BW ₂ ×S ₁ /S _th

Wherein, the BW ₁ is the second bandwidth resource, the BW ₂ is the total bandwidth resource that the controller can use, the S ₁ is the available storage space in the storage array, and the S _th is The start threshold for the GC operation.

In this method, by allocating bandwidth resources proportional to available storage space for IO operations, the accuracy of bandwidth resource allocation can be further improved, thereby improving the accuracy of storage device control.

In a possible design, the target value is the sum of the first parameter and the second parameter, the first parameter is the storage space required for the IO operation corresponding to the burst traffic, and the second parameter is The available storage space in the storage array when it is predicted that there is a burst traffic within the first time period.

In this method, the value setting of the target value can ensure that after the storage space required for the IO operation corresponding to the burst traffic is provided to the burst traffic, a certain margin of available storage space can still be maintained, so it can avoid Burst traffic will consume the available storage space, improving the stability of the storage system.

In a possible design, the storage space required for the IO operation corresponding to the burst traffic conforms to the following formula:

S ₂ =BW ₃ ×T

Wherein, the S ₂ is the storage space required for the IO operation corresponding to the burst traffic, the BW ₃ is the bandwidth resource required for transmitting the burst traffic, and the T is the duration of the burst traffic duration.

In this method, according to the bandwidth resources required to transmit the burst traffic and the duration of the burst traffic, the storage space required for the IO operation corresponding to the burst traffic can be accurately calculated, and then the storage device is controlled accordingly according to the storage space .

In a possible design, determining that the burst traffic exists includes: after the end of the second time period, determining that the burst traffic exists according to a data processing request within the second time period; wherein, the The duration of the second time period is less than the set threshold, and the end time of the second time period is before the end time of the first time period.

In this method, after the end of the second time period, based on the actual data processing requests in the second time period, it is judged whether there is burst traffic, which can realize real-time traffic identification with high accuracy.

In a possible design, determining that there is the burst traffic according to the data processing requests in the second time period includes at least one of the following: determining the total number of data processing requests in the second time period greater than the set threshold; determine that the total bandwidth of the data processing requests within the second time period is greater than the set bandwidth threshold; determine that the increment of the number of data processing requests within the second time period is greater than the set Increment threshold: determining that the growth rate of the number of data processing requests within the second time period is greater than a set growth rate threshold.

In this method, the burst traffic is characterized by a large number of data processing requests, and the number of data processing requests increases rapidly in a short period of time. Therefore, based on the multi-faceted information that can reflect the characteristics of the burst traffic, analyzing the data processing requests in the second time period can identify the burst traffic more accurately, so that the burst traffic can be more accurately identified. control.

In a possible design, predicting that there is a burst traffic in the first time period includes: using the set traffic forecasting model, according to the data processing request in the target historical time period, predicting that the burst traffic in the first time period There is a burst traffic in the segment; wherein, the traffic prediction model is obtained through machine learning and/or model training; or, according to the burst traffic prediction information from the upper layer application, it is predicted that there is a burst traffic in the first time period Burst traffic; wherein, the burst traffic prediction information is used to indicate: it is predicted that there is burst traffic within the first time period.

In this method, using the set traffic prediction model to predict the burst traffic can improve the accuracy of burst traffic prediction, and at the same time, it can adapt to the prediction of burst traffic in many different scenarios, and has good versatility. When the controller determines that the burst traffic is predicted according to the instruction from the upper-layer application, the prediction process of the burst traffic can be transferred to the upper-layer application, thereby simplifying the processing flow of the controller and increasing the processing rate.

In a possible design, the storage device is a solid state disk.

In a second aspect, the present application provides a control device for a storage device, the control device is applied to a controller in the storage device, wherein the storage device further includes a storage array; the control device includes: a prediction module, It is used to predict that there is a burst traffic within the first time period; the processing module is used to increase the bandwidth resource allocated to the GC operation before the first time period, and use the first bandwidth resource to perform the GC operation, obtaining a first available storage space in the storage array; wherein, the burst flow includes a first number of data processing requests, and the first number is greater than the number of data processing requests that the storage device can respond to per second ; The first bandwidth resource is a bandwidth resource allocated to the GC operation.

In a possible design, the processing module is further configured to: when it is determined that the burst traffic exists, allocate the first available storage space to an input/output IO operation corresponding to the burst traffic.

In a possible design, when the prediction module predicts that there is a burst traffic within the first time period, the processing module is further configured to: increase the startup threshold of the GC operation; wherein, when the When the available storage space in the storage array is less than the startup threshold, the GC operation is started.

In a possible design, when the prediction module predicts that there is a burst traffic within the first time period, the processing module is further configured to: based on the storage space required for the IO operation corresponding to the burst traffic Determine a first target time with the first bandwidth resource; when the first target time arrives, increase the start threshold of the GC operation; wherein, when the available storage space in the storage array is less than the start threshold , start the GC operation.

In a possible design, the processing module is further configured to: when it is determined that the burst traffic exists, lower the start threshold of the GC operation; or, if the predicted burst traffic is within the second target If the time has not arrived, lower the start threshold of the GC operation; wherein, the second target time is any time within the first time period.

In a possible design, the processing module is further configured to: allocate a second bandwidth resource to the current data processing request when the first available storage space reaches a target value before determining that the burst traffic exists The corresponding IO operation, and assigning the third bandwidth resource to the GC operation; wherein, the target value is greater than or equal to the storage space required for the IO operation corresponding to the burst traffic; the size of the second bandwidth resource There is a positive correlation with the size of the available storage space in the storage array, and the size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.

BW ₁ =BW ₂ ×S ₁ /S _th

S ₂ =BW ₃ ×T

In a possible design, when the processing module determines that the burst traffic exists, it is specifically configured to: after the end of the second time period, according to the data processing request in the second time period, determine that the burst traffic exists. The burst traffic; wherein, the duration of the second time period is less than a set threshold, and the end moment of the second time period is before the end moment of the first time period.

In a possible design, when the processing module determines that the burst traffic exists according to the data processing request within the second time period, it is specifically configured to perform at least one of the following: determine the second time The total number of data processing requests in the segment is greater than the set threshold; determine that the total bandwidth of the data processing requests in the second time period is greater than the set bandwidth threshold; determine the data processing requests in the second time period The increment of the quantity is greater than the set increment threshold; it is determined that the increase of the quantity of data processing requests within the second time period is greater than the set increase threshold.

In a possible design, when the prediction module predicts that there is a burst traffic in the first time period, it is specifically configured to: use the set traffic prediction model to process the request according to the data in the target historical time period, It is predicted that there is a burst traffic within the first time period; wherein, the traffic prediction model is obtained through machine learning and/or model training; or, according to the burst traffic prediction information from the upper layer application, it is predicted that the There is a burst traffic in the first time period; wherein, the burst traffic prediction information is used to indicate: it is predicted that there is a burst traffic in the first time period.

In a possible design, the storage device is a solid state disk.

In a third aspect, the present application provides a storage device, which includes a controller and a storage array; the storage array is used to store data; and the controller is used to perform the above-mentioned first aspect or the first Aspects of any possible design of the described method.

In a possible design, the storage device is a solid state disk.

In a fourth aspect, the present application provides a computing device, which includes the storage device described in the above-mentioned second aspect or any possible design of the second aspect, or, the computing device includes the above-mentioned third aspect or the storage device described in any possible design of the second aspect. Any possible design of the storage device described in the three aspects.

In a possible design, the storage device is a solid state disk.

In a fifth aspect, the present application provides a computer-readable storage medium, the computer-readable storage medium stores computer instructions, and when the computer instructions are run on a computer or a processor, the computer or the processor Execute the method described in the first aspect or any possible design of the first aspect.

In a sixth aspect, the present application provides a computer program product. When the computer program product runs on a computer or a processor, the computer or the processor executes any of the above-mentioned first aspect or any possibility of the first aspect The design of the described method.

For the beneficial effects of the above-mentioned second aspect to the sixth aspect, please refer to the description of the above-mentioned beneficial effects of the first aspect, which will not be repeated here.

Description of drawings

FIG. 1 is a schematic structural diagram of a possible application system of the solution provided by the embodiment of the present application;

FIG. 2 is a schematic diagram of a method for controlling a storage device provided in an embodiment of the present application;

FIG. 3 is a schematic flowchart of a method for controlling a storage device provided in an embodiment of the present application;

FIG. 4a is a schematic diagram of an SSD control method provided by an embodiment of the present application;

Fig. 4b is a schematic diagram of another SSD control mode provided by the embodiment of the present application;

Fig. 4c is a schematic diagram of another SSD control mode provided by the embodiment of the present application;

FIG. 5 is a schematic diagram of bandwidth resources corresponding to IO operations at different stages provided by an embodiment of the present application;

FIG. 6 is a schematic diagram of a storage device provided by an embodiment of the present application.

Detailed ways

In order to make the purpose, technical solutions, and advantages of the embodiments of the present application clearer, the embodiments of the present application will be further described in detail below in conjunction with the accompanying drawings. Among them, in the description of the embodiments of the present application, the terms "first" and "second" are used for description purposes only, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of indicated technical features . Thus, a feature defined as "first" and "second" may explicitly or implicitly include one or more of these features.

In order to facilitate understanding, descriptions of concepts related to the present application are provided as examples for reference.

1) Solid-state hard disk is a hard disk made of solid-state electronic memory chip array, also known as solid-state drive. A solid state disk generally includes at least two parts, a controller and a storage array, and the controller can be used to control the solid state disk. The storage array is used to provide the storage area of the solid-state disk, which can store data.

When flash memory is used as the storage medium for solid-state drives, the storage array usually includes dozens, hundreds or even thousands of flash memories, and each flash memory may include hundreds to thousands of flash memory blocks, and each flash memory block may include thousands of flash memory blocks. Page. Flash memory has three basic operations: read, write, and erase. Among them, the operation granularity of read operation and write operation is a flash memory page, and the operation granularity of erase operation is a flash memory block. Therefore, flash memory has remote update and asymmetric read/write erase. Latency characteristics.

In order to make better use of flash memory, solid-state drives are equipped with flash translation layer (flash translation layer, FTL) mapping tables and GC operation algorithms to ensure continuous and efficient write performance.

The FTL mapping table is an address translation table, which is used to indicate a mapping relationship between a logical address of data and a physical address. Among them, the logical address is the address generated by the system for the data before writing to the flash memory, and the physical address is the address after writing to the flash memory. After the data processing request (or user request) is sent to the solid-state hard disk, the solid-state hard disk can determine the physical address in the flash memory corresponding to the data processing request through the FTL mapping table. If the data processing request is a read request (for requesting a read operation), the data at the corresponding physical address in the flash memory is read out; if the data processing request is a write request (for requesting to perform a write operation), and the write If the operation is to write a certain data for the first time, write the data to a blank physical address in the flash memory; if the data processing request is a request to perform a write operation, and the write operation is used to update a stored data, then Mark the corresponding physical address of the data in the flash memory as invalid, and write the updated data corresponding to the data into the blank physical address in the flash memory, and then update the corresponding address mapping relationship in the FTL mapping table to complete the operation. If the data processing request is an erasing request (for requesting to perform an erasing operation), the data of the corresponding physical address in the flash memory is deleted, and the corresponding address mapping relationship in the FTL mapping table is deleted.

In this embodiment of the present application, the data processing request may be referred to as an input/output (input/output, IO) or an IO request for short.

The GC operation is to select a flash memory block to reclaim when the available storage space (blank storage space or free storage space) in the storage array is less than the startup threshold (called the GC watermark), and the specific operation is to read the data in the selected flash block. Valid data is migrated into the flash page in the available storage space, and then the content in the flash block is erased, so that the storage space corresponding to the flash block is reclaimed through GC operation. When the available storage space in the storage array is greater than or equal to the startup threshold through the GC operation, the GC operation stops execution.

2) IO (or IO request) refers to a data processing request sent to the storage device by other devices (such as a host device). Exemplarily, the IO may include at least one of a read request, a write request, and an erase request.

Taking the storage device as an example of a solid-state hard disk, according to the disk capacity of the solid-state hard disk, the IO that occurs within a period of time can be defined as two situations: regular traffic and burst traffic. Among them, the regular traffic is smaller than the disk capacity of the solid-state hard disk, which is a scenario with few data processing requests. At this time, the solid-state hard disk can process all the requests of the host in a timely manner. However, the burst traffic is greater than the disk capacity of the solid-state hard disk, which is a scenario where there are many data processing requests. At this time, the solid-state hard disk can only respond to the data processing requests in the burst traffic in batches, which may cause request response timeouts.

In the embodiment of the present application, the IO operation refers to the process in which the solid state disk performs corresponding operations according to the received IO, and is the operation performed by the solid state disk when responding to the IO. For example, the IO operation may include at least one of a read operation, a write operation, and an erase operation.

3) Bandwidth refers to the amount of data that can be transmitted per unit time. Bandwidth can be used to identify the data transmission capabilities of devices, devices or communication lines, usually expressed in bits or bits per second (bps) that can be transmitted.

Typically, the bandwidth of a storage device is fixed. In the process of controlling the execution of the IO operation or the GC operation, the controller can control the frequency of using the bandwidth to realize the effect of executing the IO operation or the GC operation under different bandwidth capabilities. Therefore, in the implementation of this application, the bandwidth resource refers to the bandwidth capability (or data transmission capability) realized by changing the frequency of using the bandwidth. Allocating bandwidth resources to IO operations or GC operations means: by configuring IO operations or GC operations to use bandwidth frequencies during operation execution, to configure corresponding bandwidth capabilities for IO operations or GC operations, so that IO operations or GC operations Execute under the corresponding bandwidth capability.

In addition, in the embodiment of the present application, the bandwidth resources of the storage device and the IOPS of the storage device (that is, the number of data processing requests that the storage device can respond to per second) can be converted to each other.

It should be understood that "at least one" in the embodiments of the present application refers to one or more, and "multiple" refers to two or more. "And/or" describes the association relationship of associated objects, indicating that there may be three types of relationships, for example, A and/or B, which can mean: A exists alone, A and B exist simultaneously, and B exists alone, where A, B can be singular or plural. The character "/" generally indicates that the contextual objects are an "or" relationship. "At least one (item) of the following" or similar expressions refer to any combination of these items, including any combination of single item(s) or plural item(s). For example, at least one item (unit) of a, b or c can represent: a, b, c, a and b, a and c, b and c, or a, b and c, wherein a, b, c Can be single or multiple.

Currently, if the IO access volume of the solid state drive is less than its disk capacity (that is, the normal traffic scenario), the solid state drive can complete the corresponding IO operations on time. Moreover, based on the current resource allocation methods for IO operations and GC operations, the bandwidth resources allocated in the SSD can also maintain a stable available storage space. However, if the IO access volume exceeds the capacity of the disk (that is, the burst traffic scenario), the solid state disk will need a long waiting time when performing the corresponding IO operation, which may cause a response timeout. Moreover, based on the current resource allocation methods for IO operations and GC operations, the bandwidth resources allocated in SSDs cannot maintain the relative stability between the storage space occupied by IO operations and the storage space released by GC, so it is impossible to maintain a stable available storage space , and there may even be a situation where the available storage space is consumed by IO operations, which will affect the storage performance and stability of the SSD.

At present, a method for controlling the resource of GC operation is to start and stop GC operation according to the requirement of IO operation, so as to achieve the effect of responding to IO first. Specifically, when the burst traffic comes, the GC operation can be stopped, so that all resources can be used for IO operations, and when the burst traffic ends, the GC operation can be started to reclaim the storage space. However, this solution only considers IO requirements without considering the actual situation of storage space, and cannot maintain a relatively stable available storage space. Moreover, when the available storage space is small and the burst traffic lasts for a long time, the available storage space may be exhausted and the system may crash. Therefore, the implementability of this scheme is very poor.

In order to improve the system stability of the solid state disk, an embodiment of the present application provides a method for controlling a storage device, wherein the storage device is capable of performing GC operations and IO operations. Optionally, the storage device may be a solid state disk. In this method, the storage device can adjust the startup threshold of the GC operation, so that the storage device can always maintain a larger available storage space in the storage device by performing the GC operation after the startup threshold is increased, and at the same time, the storage device The GC operation can be started and stopped according to the requirements of the IO operation. In this way, before the burst traffic comes, the storage device can maintain a larger available storage space by increasing the start threshold of the GC operation. When the burst traffic comes, the available storage space can be used for IO operations, thereby avoiding The IO operation will consume the available storage space, and at the same time, it can avoid the GC operation and the IO operation from competing for resources, and improve the execution efficiency of the IO operation.

In the above solution, after increasing the startup threshold of the GC operation, the storage device can always maintain a large available storage space in the normal traffic scenario, and in the burst traffic scenario, the storage device can use the available storage space when responding to the burst traffic , so as to prevent the storage system of the storage device from collapsing and improve its stability. However, in the above solution, increasing the startup threshold of GC operation will cause frequent execution of GC operation, and valid flash memory pages need to be moved during the execution of GC operation, which will increase the burden of erasing and writing of the storage device, and will cause the life of the storage device to be lost. The problem.

In view of this, the embodiment of the present application also provides another storage device control method, which is used to control the resource allocation and execution process of IO operations and GC operations of the storage device, so that the storage device can respond to a large number of IOs in a timely manner. To maintain a certain available storage space, thereby ensuring the normal working performance of the storage device, and further improving the operation stability and storage performance of the storage device.

FIG. 1 is a schematic structural diagram of a possible application system of the solution provided by the embodiment of the present application. As shown in FIG. 1 , the system includes at least a storage device, the storage device includes a controller and a storage array, the controller is used to control the storage device, and the storage array is used to provide storage space for the storage device. Optionally, the storage device may be a solid state disk.

Optionally, the system may also include a host (host) device. The host device can communicate with the storage device and can perform some control over the storage device, and the storage device can also respond to data processing requests (IO/IOs) from the host device. Optionally, the storage device may be deployed in the host device.

In the storage device, the controller may further include a statistics module, a modulation module and a GC module. Wherein, the statistical module can be used to count current traffic information, available storage space, etc., and send the counted information to the control module. The regulation module can be used to implement the control method provided by the embodiment of the present application, mainly including calculating the start threshold of GC operation under specific conditions according to current and future traffic information and information of available storage space, and allocating bandwidth resources for IO operation and GC operation , Control the execution of IO operations and GC operations, etc. The GC module is used to perform GC operations under the control of the modulation module.

In a possible solution, the controller may further include a prediction module. The prediction module can be used to predict future traffic information according to the historical traffic information of the storage device, and send the predicted future traffic information to the control module, so that the control module can obtain the future traffic information.

In another possible solution, the upper layer application in the host device can predict its future traffic information after obtaining the historical traffic information of the storage device, and send the predicted future traffic information to the controller of the storage device . Then the control module in the controller can also obtain the future traffic information from the upper layer application.

In a possible manner, the upper layer application or prediction module may convert the IOPS of the storage device into a bandwidth requirement and send it to the controller of the storage device.

In the storage device, the storage array is the physical carrier for the final storage of data. Exemplarily, the storage array may use NAND flash memory, and may include multiple flash memory arrays (flash array). Wherein, each flash memory array can be used as a NAND flash memory particle (flash memory medium particle), so the storage array is equivalent to a NAND flash memory particle group.

It should be noted that the above system architecture is only an exemplary description of the system architecture applicable to the solution of this application, and does not limit the system architecture applicable to the solution of this application. Other devices or modules may also be added to the above system architecture, or some devices or modules may be reduced or modified.

The solution provided by the present application will be introduced below in combination with specific embodiments.

FIG. 2 is a schematic diagram of a method for controlling a storage device provided by an embodiment of the present application. As shown in Figure 2, the method includes:

S201: When it is predicted that there is a burst traffic within the first time period, the controller of the storage device increases the bandwidth resources allocated to the garbage collection GC operation before the first time period; wherein the burst traffic includes A first number of data processing requests, the first number being greater than the number of data processing requests that the storage device can respond to per second.

Wherein, the storage device includes a controller and a storage array, and the storage device is capable of performing GC operations and IO operations. Optionally, the storage device is a solid state disk.

In this embodiment of the application, the number of data processing requests included in the regular traffic is less than or equal to the number of data processing requests that the storage device can respond to per second, and the number of data processing requests included in the burst traffic is greater than the number of data processing requests that the storage device can respond to per second. The number of data processing requests that can be responded to.

In a normal traffic scenario, the way the controller allocates bandwidth resources for IO operations and GC operations can be: allocate corresponding bandwidth resources for IO operations and GC operations, so that the storage device can maintain the storage required for IO operations during the running process. A relative balance is maintained between the space and the storage space reclaimed by the GC operation, for example, the size difference between the two storage spaces is within the set range. In this way, it can not only ensure that the storage device has sufficient storage space for performing IO operations, but also avoid the impact of frequent execution of GC operations on the life of the storage device as much as possible.

In the normal flow scenario, the controller performs control according to the above method, and at the same time, judges whether there is a burst flow in the future time period, so as to adjust the control method accordingly.

As an optional implementation manner, the controller may predict whether there is a burst traffic in a future time period according to data processing requests received in a historical time period. Specifically, the controller can use the set traffic forecasting model to predict that there is a burst traffic in the first time period according to the data processing request in the target historical time period; wherein, the traffic forecasting model is based on machine learning and /or obtained by model training, the first time period is a certain time period in the future. Optionally, the traffic prediction model may be a model based on a long short-term memory network (long short-term memory, LSTM).

For example, when this solution is applied to the system shown in FIG. 1 , the controller may predict that there is a burst traffic within the first time period through the prediction module.

As another optional implementation manner, the controller may predict that there is a burst traffic within the first time period according to the burst traffic prediction information from the upper-layer application; wherein the burst traffic prediction information is used for Indication: It is predicted that there is a burst traffic within the first time period. Optionally, the upper-layer application may use the above-mentioned traffic prediction model to predict that there is a burst traffic within the first time period, and indicate it to the controller.

For example, when this solution is applied to the system shown in Figure 1, the prediction of future traffic can be completed by the host device, and the controller can directly determine that there is a burst in the first time period according to the instruction of the upper layer application in the host device. flow.

In some embodiments of the present application, the information of the predicted burst traffic within the first time period may include: the expected arrival time of the burst traffic (that is, the starting moment of the first time period) , the bandwidth resources required for transmitting the burst traffic, and the duration of the burst traffic (that is, the duration of the first time period).

When the above-mentioned controller predicts that there is a burst traffic in the first time period, it may increase the bandwidth resource allocated to the GC operation on the basis of the bandwidth resource allocation in the above-mentioned normal traffic scenario. For example, the controller may increase the allocation of surplus bandwidth resources in the storage device to GC operations, thereby speeding up the reclaiming rate of the storage space.

S202: The controller uses a first bandwidth resource to execute the GC operation to obtain a first available storage space in the storage array, where the first bandwidth resource is a bandwidth resource allocated to the GC operation.

After the controller increases the bandwidth resource allocated to the GC operation, the controller may use the first bandwidth resource allocated to the GC operation to execute the GC operation to obtain the first available storage space. Wherein, the first available storage space is greater than the size of the available storage space that can be maintained in the storage device in a normal traffic scenario.

According to the above-mentioned bandwidth resource allocation method in the conventional traffic scenario, the controller only needs a part of the bandwidth resources of the storage device to complete the IO operation and GC operation, therefore, the bandwidth resources of the storage device will have surplus. Based on this, when the controller predicts that there is a burst traffic within the first time period, it can further store the Some or all of the excess bandwidth resources are increased and allocated to GC operations. The controller uses all the increased bandwidth resources allocated to the GC operation to execute the GC operation, which can speed up the execution rate of the GC operation, thereby speeding up the recovery rate of the storage space, so that the GC operation can be recovered before the arrival of the first time period. Lots of storage space. In this way, when the burst traffic arrives, more storage space can be used for IO operations in the storage device, thereby ensuring timely execution of IO operations, improving IO response speed, and avoiding too small margin of available storage space as much as possible. Improve storage device stability.

Further, in the above step S201, when it is predicted that there is a burst traffic within the first time period, the controller may increase the bandwidth resource allocated to the garbage collection GC operation before the first time period, and may also Before the first time period, the start and stop of the GC operation are controlled by increasing the start threshold of the GC operation.

In a possible manner, the controller may increase the start threshold of the GC operation when it is predicted that there is a burst traffic within the first time period. Wherein, when the available storage space in the storage array is less than the startup threshold, the GC operation is started.

In this way, by increasing the start threshold of the GC operation, the controller can obtain a larger available storage space by executing the GC operation, and it can also ensure that after the execution of the GC operation obtains a larger available storage space, the controller will The available storage space obtained can also be maintained by performing GC operations.

In another possible manner, when the controller predicts that there is a burst traffic within the first time period, based on the storage space required for the IO operation corresponding to the burst traffic and the first bandwidth resource determining a first target time; and when the first target time arrives, increasing the start threshold of the GC operation; wherein, when the available storage space in the storage array is less than the start threshold, start the GC operate.

Specifically, the controller may first calculate the storage space required for the IO operation corresponding to the burst traffic according to the predicted bandwidth resources required for transmitting the burst traffic and the duration of the burst traffic. Exemplarily, it can be calculated according to the following formula:

S ₂ =BW ₃ ×T

Wherein, S ₂ is the storage space required for the IO operation corresponding to the burst traffic, BW ₃ is the bandwidth resource required for transmitting the burst traffic, and T is the duration of the burst traffic.

Then, the controller may calculate the time required to obtain the storage space through the GC operation according to the storage space required by the IO operation corresponding to the burst traffic and the increased bandwidth resource allocated to the GC operation. Exemplarily, it can be calculated according to the following formula:

T ₁ =S ₂ /BW ₄

Wherein, t ₁ is the time required for obtaining the storage space through GC operation, S ₂ is the storage space required for the IO operation corresponding to the burst traffic, and BW ₄ is the additional allocation of the first resource to the Bandwidth resources for GC operations.

After the above information is determined, the controller may calculate and determine the first target moment. Specifically, the controller may calculate the time length between the first target moment and the moment when it is predicted that there is a burst traffic within the first time period according to the following formula:

T ₂ =T ₃ -T ₁ -δ

Among them, _T2 is the time length between the first target moment and the moment when the burst traffic is predicted in the first time period, _T3 is the start moment of the first time period and the time when the first time period is predicted The length of time between the moments of burst traffic in the time period, T ₁ is the time required to obtain the storage space required for the IO operation corresponding to the burst traffic through GC operations, δ can be the set duration, and δ is greater than or equal to 0 And less than or equal to T ₃ -T ₁ .

After calculating T ₂ , the controller delays by T ₂ the time when the burst traffic is predicted to exist in the first time period, and the corresponding time is the first target time. Then, when the first target time arrives, the start threshold of the GC operation can be increased.

In this way, the controller can determine the time T ₁ required to obtain the storage space that can be used to handle the burst traffic through the GC operation according to the additional bandwidth resources allocated to the GC operation and the bandwidth resources required to transmit the burst traffic, and then Adjusting the GC threshold from the time when the estimated arrival time of the burst traffic is ahead of T ₁ can not only minimize unnecessary GC operations, reduce the impact on the life of the storage device, but also obtain a greater Storage space for use when burst traffic arrives. Furthermore, adding a certain time margin δ on the basis of advancing T ₁ can further ensure that the storage space obtained through the GC operation is sufficient for the arrival of burst traffic.

In the above two manners, when the controller increases the start threshold of the GC operation, the start threshold of the GC operation may be adjusted to the first threshold. Wherein, the first threshold is greater than or equal to a target value, and the target value is greater than or equal to the storage space required by the IO operation corresponding to the burst traffic.

Exemplarily, the target value may be the sum of the first parameter and the second parameter, the first parameter is the storage space required for the IO operation corresponding to the burst traffic, and the second parameter is the predicted The available storage space in the storage array when there is a burst of traffic in the first time period. The first threshold may be equal to the target value or slightly greater than the target value. That is to say, the target value can satisfy the following formula: S=S ₂ +S ₃ , where S is the target value, S ₂ is the storage space required for the IO operation corresponding to the burst traffic, and S ₃ is the predicted The available storage space in the storage array when there is a burst of traffic in the first time period. The first threshold may satisfy the following formula: N ₁ ≥ S, where N ₁ is the first threshold.

Further, after the controller allocates bandwidth resources for GC operation based on the above method and increases the start threshold of GC operation, in the process of executing GC operation according to step S202, it can judge whether there is burst traffic according to real-time traffic information (that is, Whether the burst traffic is coming), so as to carry out further control.

Wherein, when the controller judges whether there is a burst traffic according to the real-time traffic information, the following method may be adopted: after the second time period ends, according to the data processing request in the second time period, it is determined that the burst traffic exists; Wherein, the duration of the second time period is less than a set threshold, and the end time of the second time period is before the end time of the first time period. Exemplarily, the set threshold may be, for example, a short duration of millisecond level or second level.

Specifically, when the controller determines that the data processing request within the second time period satisfies at least one of the following conditions 1 to 4, it may determine that there is a burst traffic.

Condition 1. The total number of data processing requests within the second time period is greater than a set number threshold.

Condition 2. The total bandwidth of the data processing requests within the second time period is greater than the set bandwidth threshold.

Condition 3. The increment of the number of data processing requests within the second time period is greater than a set increment threshold.

Condition 4: The growth rate of the number of data processing requests within the second time period is greater than a set growth rate threshold.

Wherein, the second time period in each of the above conditions is a relatively short time period, and when the data processing requests in the second time period satisfy condition 1, it means that there are a large number of data processing requests in a short period of time, and it can be considered that there is a burst send traffic. When the data processing request in the second time period satisfies condition 2, it means that the data volume of the data processing request in a short period of time is large, and it can be considered that there is a burst traffic. When the data processing requests in the second time period satisfy condition 3 or condition 4, it means that the data processing requests in a short period of time increase rapidly, and it can be considered that there is a burst traffic.

It should be noted that the embodiments of the present application include but are not limited to the above method for judging whether there is burst traffic, and the controller may also use other methods for judging, which will not be listed here.

In the embodiment of the present application, when the controller judges in real time whether there is a burst traffic, the judgment results may be obtained in various situations, and the corresponding control methods adopted are also different. At the same time, when the size of the available storage space in the storage array changes, the control mode adopted by the controller also changes accordingly. Therefore, the controller needs to flexibly adjust the control mode of GC operation and IO operation according to the change of the actual situation, so as to obtain a better control effect.

The following describes in detail in conjunction with the following cases 1 to 3.

Case 1. Before the arrival of burst traffic, the available storage space in the storage array reaches the target value, and the controller reallocates the bandwidth resources used during the execution of IO operations and GC operations.

It should be noted that, in the embodiment of the present application, when the controller determines that there is a burst traffic according to the above manner, it indicates that the burst traffic is coming.

In the embodiment of the present application, the target value is greater than or equal to the storage space required by the IO operation corresponding to the burst traffic. Therefore, when the available storage space in the storage array reaches the target value, it indicates that the first available storage space obtained through the GC operation is sufficient for the IO operation corresponding to the burst traffic. If it is not determined that there is a burst of traffic at this time, that is, the burst of traffic has not yet arrived, the execution speed of the IO operation can be appropriately reduced so that it can maintain the first available storage space currently obtained, thereby avoiding excessive GC operations. Effects on storage device lifetime.

Specifically, the controller may allocate the second bandwidth resource to the IO operation corresponding to the current data processing request when determining that the first available storage space reaches the target value before determining that the burst traffic exists, and allocate the third bandwidth resource to Assigned to GC operations. Wherein, the size of the second bandwidth resource has a positive correlation with the size of the available storage space in the storage array, and the size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.

Based on this method, when the available storage space is large, the controller can allocate more bandwidth resources for IO operations and less bandwidth resources for GC operations, which can speed up the consumption of available storage space and slow down the consumption of available storage space. Recycling, so as to properly consume the available storage space, avoid excessive storage space, and improve the speed of IO operations. When the available storage space is small, the controller can allocate less bandwidth resources for IO operations and more bandwidth resources for GC operations, which can speed up the recovery of available storage space and slow down the consumption of available storage space, thereby avoiding unavailable The storage space is too small. By flexibly allocating bandwidth resources for IO operations and GC operations in this way, the size of the available storage space can be kept stable until burst traffic arrives.

Exemplarily, the above-mentioned second bandwidth resource conforms to the following formula:

BW ₁ =BW ₂ ×S ₁ /S _th

Wherein, BW ₁ is the second bandwidth resource, BW ₂ is the total bandwidth resource that the controller can use, S ₁ is the available storage space in the storage array, and the S _th is the start threshold of GC operation.

The above-mentioned third bandwidth resource may be a part or all of the bandwidth resource obtained by subtracting the second bandwidth resource from the total bandwidth resource available to the controller.

In case 2, when the burst traffic comes, the controller lowers the start threshold of the GC operation.

During the process of judging whether there is a burst flow in real time, the controller lowers the start threshold of the GC operation when it is determined that there is a burst flow. At the same time, the controller can allocate the first available storage space obtained by executing the GC operation to the IO operation corresponding to the burst traffic, so as to ensure a timely response to the burst traffic.

Among them, after the controller lowers the start threshold of the GC operation, the GC operation stops, and more bandwidth resources can be used for the IO operation, and the IO operation can respond to the burst traffic in a timely and rapid manner, thereby reducing or avoiding the IO response timeout The problem.

When the controller lowers the GC operation start threshold, the GC operation start threshold may be adjusted to the second threshold. Wherein, the second threshold may be the start threshold (N ₀ ) of the GC operation when a burst traffic is predicted, or the second threshold is less than or equal to the target value (S).

Situation 3: If the predicted burst traffic does not arrive at the second target time, lower the start threshold of the GC operation; wherein, the second target time is any time within the first time period .

When the predicted burst traffic does not exist, if a high GC operation start threshold is maintained all the time, the life span of the storage device will be affected due to excessive and unnecessary GC operations. Therefore, when it is determined that the predicted burst traffic has not come, it is necessary to lower the start threshold of the GC operation in time. Since the predicted burst traffic exists for the first period of time, if the arrival of burst traffic cannot be determined within the first time period, the controller can determine that the prediction is wrong and there is no burst traffic in fact, and lower the GC The start threshold for the action.

Specifically, if the controller determines that the predicted burst flow does not arrive at the second target moment, it determines that the burst flow does not exist, and may lower the start threshold of the GC operation. Optionally, the second target moment may be the start moment of the first time period. Wherein, when the controller lowers the start threshold of the GC operation, the start threshold of the GC operation may be adjusted to the second threshold described in the above case 2.

In the above embodiment, the controller in the storage device predicts the arrival of burst traffic in advance, and at the same time, considering that the bandwidth resources required by the conventional traffic are small and there are surplus bandwidth resources available in the storage device, when the burst traffic is predicted When , increase the bandwidth resources allocated to GC operations, thereby speeding up the recovery rate of storage space, and preparing available storage space for burst traffic in advance. Then when the burst traffic comes, the controller can provide the available storage space prepared in advance to the IO operation corresponding to the burst traffic, so as to ensure the speed of IO response, improve the user experience, and avoid excessive consumption of the available storage space , improving the stability of the operation of the storage device.

Taking the storage device as an SSD as an example, the solution provided by the embodiment of the present application will be introduced below.

FIG. 3 is a schematic flowchart of a method for controlling a storage device provided by an embodiment of the present application. When the storage device is an SSD, the process includes:

S301: In a normal traffic scenario, the controller controls the SSD according to a normal policy, and obtains prediction information of burst traffic.

Wherein, the conventional strategy is to allocate corresponding bandwidth resources to the GC operation according to the executed IO operations, so that the storage space reclaimed by the GC operation is relatively balanced with the storage space required by the IO operations corresponding to the conventional traffic.

Exemplarily, referring to FIG. 4 a , initially, the controller of the SSD may adopt flow control policy 0 for regulation, and receive prediction information of burst traffic. Wherein, flow control strategy 0 is to control the startup threshold of GC operation as an initial value N ₀ . At this time, the bandwidth resources for IO operations are similar to those for GC operations, and both are smaller. There are many idle bandwidth resources in SSDs that are not being used.

The prediction information of the burst traffic includes the expected arrival time of the burst traffic, the duration, the bandwidth resource required for transmitting the burst traffic, and the like. The prediction information of the burst traffic can be obtained by the controller itself, or determined according to the instruction of the upper layer application.

S302: After the controller obtains the prediction information of the burst traffic, increase the bandwidth resource allocated to the GC operation, and increase the start threshold of the GC operation.

Exemplarily, referring to FIG. 4 b , the controller adjusts the adopted flow control policy 0 to the flow control policy 1 after determining that the burst traffic is predicted according to the prediction information of the burst traffic. Among them, flow control strategy 1 is to adjust the start threshold of GC operation to the first threshold N ₁ , and at the same time, increase and allocate idle bandwidth resources in SSD to GC operation until the available storage space reaches the target value S.

Wherein, the target value S=bandwidth resources required for transmitting burst traffic×burst traffic duration+available storage space in the storage array when burst traffic is predicted, and N ₁ is a value slightly larger than S.

Exemplarily, as shown in the curve corresponding to the "stage of preparing available storage space" in Figure 5, when a burst traffic is predicted, the controller adopts flow control strategy 1, and the bandwidth resource used by the IO operation is based on the size of the actual IO traffic. However, since the space required for IO operations is small and the speed of GC operations is accelerating, the size of the available storage space is gradually increasing.

S303: When the controller determines that the available storage space reaches the target value, it allocates bandwidth resources that have a positive correlation with the current available storage space for IO operations, and allocates bandwidth resources that have a negative correlation with the current available storage space for GC operations.

Exemplarily, referring to FIG. 4 b , after the controller determines that the available storage space reaches the target value S, the controller adjusts the adopted flow control strategy 1 to flow control strategy 2 . Among them, flow control strategy 2 is to allocate bandwidth resources proportional to the available storage space to IO operations, and the remaining bandwidth resources can be allocated to GC operations.

Wherein, the bandwidth resource corresponding to the IO operation=total bandwidth resource×available storage space/N ₁ . The maximum available storage space is N ₁ , and all bandwidth resources are used for IO operations at this time.

Exemplarily, as shown in the curve corresponding to the stage of "maintaining available storage space" in Figure 5, when the controller adopts flow control strategy 2, the bandwidth resources used by IO operations are proportional to the available storage space. During this process, it can be maintained as much as possible The available storage space is at the level of S or so.

S304: The controller judges whether the burst traffic is coming according to the actual IO situation, and if so, executes step S305; otherwise, executes step S306.

Exemplarily, if the controller determines that within 3 consecutive seconds, the number of IOs per second increases by more than 50% compared with the previous second, it may determine that the burst traffic is coming.

S305: the controller lowers the start threshold of the GC operation.

Exemplarily, referring to FIG. 4 c , the controller adjusts the adopted flow control strategy 2 to flow control strategy 3 when determining that a burst traffic is coming. Among them, the flow control strategy 3 is to adjust the start threshold of the GC operation to the initial value N ₀ , so as to stop the GC operation and use the bandwidth resource for the IO operation.

Exemplarily, as shown in the curve corresponding to the "stage of using available storage space" in FIG. 5, when the controller adopts the flow control strategy 3, the GC operation stops, and the IO operation can use the total bandwidth resource.

S306: The controller maintains the current control mode, or switches to control the SSD according to a conventional strategy when determining the expected end time of the arrival of the burst traffic.

Exemplarily, when the controller does not determine that the burst traffic is coming, it may maintain the flow control policy 2, or, when determining the expected end time of the arrival of the burst traffic, adjust the adopted traffic control policy 2 to the traffic control policy 3.

Wherein, the expected end time of the burst traffic is the corresponding time after the expected arrival time of the burst traffic is delayed by the duration of the burst traffic.

It should be understood that the above example is only an exemplary description of an implementation manner in a possible scenario of the solution of the present application, and does not cover all possible scenarios or situations. For the specific implementation manner of each step in the above example, reference may be made to the introduction in the above embodiment, and details are not described here again.

In the above example, after predicting the burst traffic, the controller can quickly free up the storage space required for the IO operation corresponding to the burst traffic, and lower the startup threshold of the GC operation when the burst traffic comes, so as to maximize the IO response Ability to absorb burst traffic, improve IO response speed and SSD stability. During this process, only when the allocation of bandwidth resources is increased for the GC operation to quickly reclaim the storage space will have an impact on the life of the SSD, but the duration is short, so the impact on the life can be reduced.

Based on the above embodiments and the same idea, an embodiment of the present application further provides a control device for a storage device, the control device is applied to a controller in the storage device, and the storage device further includes a storage array. As shown in FIG. 6 , the control device 600 includes: a predicting module 601, configured to predict that there is a burst traffic within the first time period; a processing module 602, used to increase the allocation to Bandwidth resources for GC operations, and using the first bandwidth resources to execute the GC operations to obtain a first available storage space in the storage array; wherein, the burst traffic includes a first number of data processing requests, and the second A number is greater than the number of data processing requests that the storage device can respond to per second; the first bandwidth resource is a bandwidth resource allocated to the GC operation.

In a possible design, the processing module 602 is further configured to: when it is determined that the burst traffic exists, allocate the first available storage space to an input/output IO operation corresponding to the burst traffic.

In a possible design, when the prediction module 601 predicts that there is a burst traffic within the first time period, the processing module 602 is further configured to: increase the startup threshold of the GC operation; wherein, when When the available storage space in the storage array is less than the start threshold, start the GC operation.

In a possible design, when the prediction module 601 predicts that there is a burst traffic within the first time period, the processing module 602 is further configured to: based on the IO operation required by the burst traffic storage space and the first bandwidth resource determine a first target time; when the first target time arrives, increase the start threshold of the GC operation; wherein, when the available storage space in the storage array is less than the When the threshold is started, the GC operation is started.

In a possible design, the processing module 602 is further configured to: when it is determined that the burst traffic exists, lower the start threshold of the GC operation; or, if the predicted burst traffic is in the second If the target time has not arrived, lower the start threshold of the GC operation; wherein, the second target time is any time within the first time period.

In a possible design, the processing module 602 is further configured to: before it is determined that the burst traffic exists, when the first available storage space reaches a target value, allocate the second bandwidth resource to the current data processing Requesting a corresponding IO operation, and allocating a third bandwidth resource to the GC operation; wherein, the target value is greater than or equal to the storage space required for the IO operation corresponding to the burst traffic; the second bandwidth resource The size has a positive correlation with the size of the available storage space in the storage array, and the size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.

BW ₁ =BW ₂ ×S ₁ /S _th

S ₂ =BW ₃ ×T

In a possible design, when the processing module 602 determines that the burst traffic exists, it is specifically configured to: after the end of the second time period, according to the data processing request in the second time period, determine that there is The burst traffic; wherein, the duration of the second time period is less than a set threshold, and the end moment of the second time period is before the end moment of the first time period.

In a possible design, when the processing module 602 determines that the burst traffic exists according to the data processing request within the second time period, it is specifically configured to perform at least one of the following: determine the second The total number of data processing requests in the time period is greater than the set number threshold; determine that the total bandwidth of the data processing requests in the second time period is greater than the set bandwidth threshold; determine the data processing in the second time period The increment of the quantity of requests is greater than a set increment threshold; it is determined that the increment of the quantity of data processing requests within the second time period is greater than the set increment threshold.

In a possible design, when the prediction module 601 predicts that there is a burst traffic in the first time period, it is specifically configured to: use the set traffic prediction model to process the request according to the data in the target historical time period , it is predicted that there is a burst traffic within the first time period; wherein, the traffic prediction model is obtained through machine learning and/or model training; or, according to the burst traffic prediction information from the upper-layer application, it is predicted that There is a burst traffic within the first time period; wherein, the burst traffic prediction information is used to indicate: it is predicted that there is a burst traffic within the first time period.

In a possible design, the storage device is a solid state disk.

The division of modules in the embodiments of the present application is schematic, and is only a logical function division. There may be other division methods in actual implementation. In addition, each functional module in each embodiment of the present application can be integrated into a processing In the controller, it can also be physically present separately, or two or more modules can be integrated into one module. The above-mentioned integrated modules can be implemented in the form of hardware or in the form of software function modules.

Only one or more of the various modules in FIG. 6 may be implemented in software, hardware, firmware or a combination thereof. The software or firmware includes but is not limited to computer program instructions or codes, and can be executed by a hardware processor. The hardware includes but not limited to various integrated circuits, such as central processing unit (CPU), digital signal processor (DSP), field programmable gate array (FPGA) or application specific integrated circuit (ASIC).

Based on the above embodiments and the same idea, the embodiment of the present application also provides a storage device, the storage device includes a controller and a storage array; the storage array is used to store data; the controller is used to control the storage device The method for controlling the storage device provided in the foregoing embodiments is executed.

In a possible design, the storage device is a solid state disk.

Based on the above embodiments and the same idea, the embodiment of the present application further provides a computing device, the computing device includes the storage device provided in the above embodiment.

In a possible design, the storage device is a solid state disk.

Based on the above embodiments and the same idea, the embodiment of the present application also provides a computer-readable storage medium, the computer-readable storage medium stores computer instructions, and when the computer instructions are run on a computer or a processor, the The computer or the processor executes the method for controlling a storage device provided in the foregoing embodiments.

Based on the above embodiments and the same idea, the embodiment of the present application also provides a computer program product, when the computer program product runs on a computer or a processor, the computer or the processor executes the computer program provided in the above embodiment. A control method for a storage device.

The methods provided in the embodiments of the present application may be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using software, it may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on the computer, the processes or functions according to the embodiments of the present invention will be generated in whole or in part. The computer may be a general purpose computer, a special purpose computer, a computer network, network equipment, user equipment or other programmable devices. The computer instructions may be stored in or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website, computer, server or data center Transmission to another website site, computer, server, or data center by wired (such as coaxial cable, optical fiber, digital subscriber line (DSL) or wireless (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device such as a server, data center, etc. integrated with one or more available media. The available medium can be a magnetic medium (for example, a floppy disk, a hard disk, a magnetic tape), optical media (for example, digital video disc (digital video disc, DVD for short), or semiconductor media (for example, SSD), etc.

Apparently, those skilled in the art can make various changes and modifications to the present application without departing from the scope of the present application. In this way, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application is also intended to include these modifications and variations.

Claims

A method for controlling a storage device, the storage device comprising a controller and a storage array, the method being applied to the controller, wherein the method includes:

When it is predicted that there is a burst traffic within the first time period, before the first time period, increase the bandwidth resource allocated to the garbage collection GC operation; wherein the burst traffic includes a first number of data processing requests , the first number is greater than the number of data processing requests that the storage device can respond to per second;

Executing the GC operation by using the first bandwidth resource to obtain a first available storage space in the storage array, where the first bandwidth resource is a bandwidth resource allocated to the GC operation.
The method according to claim 1, further comprising:

When it is determined that the burst traffic exists, the first available storage space is allocated to an input/output IO operation corresponding to the burst traffic.
The method according to claim 1 or 2, wherein when it is predicted that there is a burst traffic within the first time period, the method further comprises:

Increase the startup threshold of the GC operation; wherein, when the available storage space in the storage array is smaller than the startup threshold, the GC operation is started.
The method according to claim 1 or 2, wherein when it is predicted that there is a burst traffic within the first time period, the method further comprises:

determining a first target moment based on the storage space required for the IO operation corresponding to the burst traffic and the first bandwidth resource;

When the first target time arrives, the start threshold of the GC operation is increased; wherein, when the available storage space in the storage array is smaller than the start threshold, the GC operation is started.
The method according to any one of claims 1-4, characterized in that the method further comprises:

When it is determined that the burst traffic exists, lowering the start threshold of the GC operation; or

If the predicted burst traffic does not arrive at the second target time, lowering the start threshold of the GC operation; wherein the second target time is any time within the first time period.
The method according to any one of claims 1-5, characterized in that the method further comprises:

Before determining that there is the burst traffic, when the first available storage space reaches a target value, allocate a second bandwidth resource to the IO operation corresponding to the current data processing request, and allocate a third bandwidth resource to the GC operate;

Wherein, the target value is greater than or equal to the storage space required by the IO operation corresponding to the burst traffic; the size of the second bandwidth resource has a positive correlation with the size of the available storage space in the storage array, so The size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.
The method according to claim 6, wherein the second bandwidth resource complies with the following formula:

BW 1 =BW 2 ×S 1 /S th

Wherein, the BW 1 is the second bandwidth resource, the BW 2 is the total bandwidth resource that the controller can use, the S 1 is the available storage space in the storage array, and the S th is The start threshold for the GC operation.
The method according to claim 6 or 7, wherein the target value is the sum of the first parameter and the second parameter, and the first parameter is the storage space required for the IO operation corresponding to the burst traffic , the second parameter is the available storage space in the storage array when it is predicted that there is a burst traffic within the first time period.
The method according to any one of claims 6 to 8, wherein the storage space required for the IO operation corresponding to the burst traffic meets the following formula:

S 2 =BW 3 ×T

Wherein, the S 2 is the storage space required for the IO operation corresponding to the burst traffic, the BW 3 is the bandwidth resource required for transmitting the burst traffic, and the T is the duration of the burst traffic duration.
The method according to any one of claims 6-9, wherein determining that there is a burst traffic comprises:

After the second time period ends, according to the data processing request in the second time period, it is determined that the burst traffic exists; wherein, the duration of the second time period is less than a set threshold, and the second time period The end moment of is before the end moment of the first time period.
The method according to claim 10, wherein, according to the data processing request within the second time period, determining that there is the burst traffic includes at least one of the following:

determining that the total number of data processing requests within the second time period is greater than a set number threshold;

determining that the total bandwidth of data processing requests within the second time period is greater than a set bandwidth threshold;

determining that the increment of the number of data processing requests within the second time period is greater than a set increment threshold;

It is determined that the growth rate of the number of data processing requests within the second time period is greater than a set growth rate threshold.
The method according to any one of claims 1-11, wherein predicting that there is a burst traffic within the first time period includes:

Using the set traffic prediction model, according to the data processing requests in the target historical time period, it is predicted that there is a burst traffic in the first time period; wherein, the traffic prediction model is through machine learning and/or model training obtained; or

According to the burst traffic prediction information from the upper-layer application, it is predicted that there is a burst traffic within the first time period; wherein, the burst traffic prediction information is used to indicate: it is predicted that there is a burst traffic within the first time period burst traffic.
The method according to any one of claims 1-12, wherein the storage device is a solid state disk.
A control device for a storage device, the control device is applied to a controller in the storage device, wherein the storage device further includes a storage array; it is characterized in that the control device includes:

A prediction module, configured to predict that there is a burst traffic in the first time period;

A processing module, configured to increase bandwidth resources allocated to GC operations before the first time period, and use the first bandwidth resources to perform the GC operations to obtain a first available storage space in the storage array; wherein, The burst traffic includes a first number of data processing requests, and the first number is greater than the number of data processing requests that the storage device can respond to per second; the first bandwidth resource is allocated to the GC operation bandwidth resources.
The control device according to claim 14, wherein the processing module is further used for:

When it is determined that the burst traffic exists, the first available storage space is allocated to an input/output IO operation corresponding to the burst traffic.
The control device according to claim 14 or 15, wherein when the prediction module predicts that there is a burst traffic within the first time period, the processing module is further configured to:

Increase the startup threshold of the GC operation; wherein, when the available storage space in the storage array is smaller than the startup threshold, the GC operation is started.
The control device according to claim 14 or 15, wherein when the prediction module predicts that there is a burst traffic within the first time period, the processing module is further configured to:

determining a first target moment based on the storage space required for the IO operation corresponding to the burst traffic and the first bandwidth resource;

When the first target time arrives, the start threshold of the GC operation is increased; wherein, when the available storage space in the storage array is smaller than the start threshold, the GC operation is started.
The control device according to any one of claims 14-17, wherein the processing module is further used for:

When it is determined that the burst traffic exists, lowering the start threshold of the GC operation; or

If the predicted burst traffic does not arrive at the second target time, lowering the start threshold of the GC operation; wherein the second target time is any time within the first time period.
The control device according to any one of claims 14-18, wherein the processing module is further used for:

Before determining that there is the burst traffic, when the first available storage space reaches a target value, allocate a second bandwidth resource to the IO operation corresponding to the current data processing request, and allocate a third bandwidth resource to the GC operate;

Wherein, the target value is greater than or equal to the storage space required by the IO operation corresponding to the burst traffic; the size of the second bandwidth resource has a positive correlation with the size of the available storage space in the storage array, so The size of the third bandwidth resource has a negative correlation with the size of the available storage space in the storage array.
The control device according to claim 19, wherein the second bandwidth resource complies with the following formula:

BW 1 =BW 2 ×S 1 /S th

Wherein, the BW 1 is the second bandwidth resource, the BW 2 is the total bandwidth resource that the controller can use, the S 1 is the available storage space in the storage array, and the S th is The start threshold for the GC operation.
The control device according to claim 19 or 20, wherein the target value is the sum of the first parameter and the second parameter, and the first parameter is the storage required for the IO operation corresponding to the burst traffic. space, and the second parameter is the available storage space in the storage array when it is predicted that there is a burst traffic within the first time period.
The control device according to any one of claims 19-21, wherein the storage space required for the IO operation corresponding to the burst traffic conforms to the following formula:

S 2 =BW 3 ×T

Wherein, the S 2 is the storage space required for the IO operation corresponding to the burst traffic, the BW 3 is the bandwidth resource required for transmitting the burst traffic, and the T is the duration of the burst traffic duration.
The control device according to any one of claims 19-22, characterized in that, when the processing module determines that the burst traffic exists, it is specifically used for:

After the second time period ends, according to the data processing request in the second time period, it is determined that the burst traffic exists; wherein, the duration of the second time period is less than a set threshold, and the second time period The end moment of is before the end moment of the first time period.
The control device according to claim 23, wherein the processing module is specifically configured to perform at least one of the following when determining that the burst traffic exists according to the data processing request within the second time period:

determining that the total number of data processing requests within the second time period is greater than a set number threshold;

determining that the total bandwidth of data processing requests within the second time period is greater than a set bandwidth threshold;

determining that the increment of the number of data processing requests within the second time period is greater than a set increment threshold;

It is determined that the growth rate of the number of data processing requests within the second time period is greater than a set growth rate threshold.
The control device according to any one of claims 14-23, wherein the prediction module is specifically used for:

Using the set traffic prediction model, according to the data processing requests in the target historical time period, it is predicted that there is a burst traffic in the first time period; wherein, the traffic prediction model is through machine learning and/or model training obtained; or

According to the burst traffic prediction information from the upper-layer application, it is predicted that there is a burst traffic within the first time period; wherein, the burst traffic prediction information is used to indicate: it is predicted that there is a burst traffic within the first time period burst traffic.
The control device according to any one of claims 14-25, wherein the storage device is a solid-state hard disk.
A storage device, characterized in that the storage device includes a controller and a storage array;

The storage array is used to store data;

The controller is configured to execute the method according to any one of claims 1-13 on the storage device.
The storage device according to claim 27, wherein the storage device is a solid-state hard disk.
A computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, and when the computer instructions are run on a computer or a processor, the computer or the processor executes the The method described in any one of claims 1-13.
A computer program product, characterized in that, when the computer program product is run on a computer or a processor, the computer or the processor is made to execute the method according to any one of claims 1-13.