A method and system for adaptively adjusting data reconstruction speed

Abstract

The present invention relates to the field of data reconstruction technology, and provides a method and system for adaptively adjusting data reconstruction speed. The method of the present invention comprises: obtaining business information of a storage node that needs to be restored and information about the consumption of hardware resources by the storage node; judging the consumption state of hardware resources as busy or idle according to the business information of the storage node that needs to be restored and information about the consumption of hardware resources by the storage node; and adjusting the data reconstruction speed according to the consumption state of hardware resources. The method and system for adaptively adjusting data reconstruction speed according to an exemplary embodiment of the present invention can improve the accuracy and independence of data reconstruction flow control, improve the efficiency of data reconstruction without affecting the host business, and improve the automation and real-time response efficiency of data reconstruction.

Description

Technical Field

[0001] The present invention relates to the technical field of data reconstruction, and in particular to a method and system for adaptively adjusting data reconstruction speed. Background Art

[0002] In a distributed storage cluster, when a storage node in the cluster fails or recovers, the cluster will be triggered to reconstruct data. If there are many storage nodes involved in data reconstruction or data that needs to be reconstructed in the cluster, data reconstruction will consume a large amount of cluster resources such as hard disks, CPUs, and networks. In order to ensure that the cluster can normally process IO read and write requests sent by clients during data reconstruction, flow control restrictions need to be imposed on data reconstruction in the cluster to reduce the impact of data reconstruction on host IO.

[0003] At present, in the prior art, the main way to limit data reconstruction is to limit the data reconstruction bandwidth or limit the concurrency of data reconstruction task execution. For the method of limiting the data reconstruction bandwidth, a reconstruction bandwidth threshold is directly set for the cluster. When the cluster reconstruction speed reaches the threshold, the execution of the data reconstruction task will be limited. In practical applications, the use of limiting the data reconstruction bandwidth as a flow control method requires cluster operation and maintenance personnel to manually set and adjust the relevant thresholds according to the cluster scale and actual scenarios. In addition, in order to prevent the setting of incorrect thresholds from having a greater impact on the host business, the range of thresholds that can be set provided by the cluster management side is limited. In fact, clusters of different scales or different hardware components have different requirements for these thresholds. In this way, it is difficult to adjust the cluster data reconstruction to a more reasonable range by operation and maintenance settings alone. Limiting the concurrency of data reconstruction task execution can strictly limit the resource consumption of data reconstruction business. By limiting the concurrency of data reconstruction task execution, the actual effect is different for nodes with different hardware conditions, and the corresponding reconstruction speed cannot be calculated according to the number of concurrent tasks. The entire flow control process is inaccurate. In addition, the above restrictions are for the entire cluster. In fact, different storage nodes are responsible for different host and data reconstruction services. It is impossible to make reasonable adjustments to the data reconstruction service simply by adjusting a global threshold.

[0004] Therefore, how to provide a more accurate data reconstruction flow control restriction method has become a technical problem that needs to be solved urgently. Summary of the invention

[0005] In view of this, in order to improve the accuracy and efficiency of data reconstruction flow control, the present invention provides a method for adaptively adjusting data reconstruction speed.

[0006] In one aspect, the present invention provides a method for adaptively adjusting data reconstruction speed, comprising:

[0007] Step S1: Obtaining service information of a storage node that needs to be restored and information about hardware resource consumption of the storage node;

[0008] Step S2: judging the consumption state of the hardware resources as busy or idle based on the service information of the storage node that needs to be restored and the consumption information of the hardware resources by the storage node;

[0009] Step S3: adjusting the data reconstruction speed according to the consumption status of hardware resources.

[0010] Furthermore, in step S1 of the method for adaptively adjusting data reconstruction speed of the present invention, the service information of the storage node includes: the bandwidth of the host service carried by each disk and the bandwidth of the data reconstruction service.

[0011] Furthermore, in step S1 of the method for adaptively adjusting data reconstruction speed of the present invention, the consumption information of hardware resources by the storage node includes: the CPU cores occupied by each disk.

[0012] Furthermore, step S2 of the method for adaptively adjusting data reconstruction speed of the present invention comprises:

[0013] The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy.

[0014] Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing;

[0015] Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, determine that the storage node network is busy;

[0016] When one of the CPU cores occupied by the disk, disk reading and writing, or storage node network is judged to be busy, the consumption status of the hardware resource is judged to be busy. When none of the three items of CPU cores occupied by the disk, disk reading and writing, and storage node network are judged to be busy, the consumption status of the hardware resource is judged to be idle.

[0017] Furthermore, step S3 of the method for adaptively adjusting data reconstruction speed of the present invention comprises:

[0018] When the consumption status of hardware resources is busy, determine whether there is a host business;

[0019] When there is no host business, the current round of data reconstruction speed adjustment ends;

[0020] When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

[0021] Furthermore, step S3 of the method for adaptively adjusting data reconstruction speed of the present invention further includes:

[0022] When the consumption status of hardware resources is idle, the data recovery speed is improved and it is determined whether there is host business;

[0023] When there is no host business, the token bucket model is used to generate the theoretical data reconstruction speed. The actual data reconstruction speed is compared with the theoretical data reconstruction speed. When the two are consistent, the data recovery speed is further improved. When the two are inconsistent, the current round of data reconstruction speed adjustment is terminated.

[0024] When there is a host business, monitor the processing speed of the host business. When the processing speed of the host business does not decrease, generate the theoretical data reconstruction speed through the token bucket model, compare the actual data reconstruction speed with the theoretical data reconstruction speed, and when the two are consistent, continue to improve the data recovery speed. When the two are inconsistent, end this round of data reconstruction speed adjustment; when the processing speed of the host business decreases, end this round of data reconstruction speed adjustment.

[0025] In another aspect, the present invention provides a system for adaptively adjusting data reconstruction speed, comprising:

[0026] A storage node information acquisition module is used to acquire the service information of the storage node that needs to be restored and the consumption information of the hardware resources of the storage node;

[0027] A hardware resource consumption state judgment module is used to judge the consumption state of the hardware resource as busy or idle based on the service information of the storage node to be restored and the consumption information of the hardware resource by the storage node;

[0028] The data reconstruction speed adjustment module is used to adjust the data reconstruction speed according to the consumption status of hardware resources.

[0029] Furthermore, in the data reconstruction speed adaptive adjustment system of the present invention, the storage node information acquisition module is specifically used to obtain the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, and obtain the CPU cores occupied by each disk.

[0030] Furthermore, in the data reconstruction speed adaptive adjustment system of the present invention, the hardware resource consumption state judgment module is specifically used for:

[0031] The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy.

[0032] Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing;

[0033] Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, determine that the storage node network is busy;

[0034] When one of the CPU cores occupied by the disk, disk reading and writing, or storage node network is judged to be busy, the consumption status of the hardware resource is judged to be busy. When none of the three items of CPU cores occupied by the disk, disk reading and writing, and storage node network are judged to be busy, the consumption status of the hardware resource is judged to be idle.

[0035] Furthermore, in the data reconstruction speed adaptive adjustment system of the present invention, the data reconstruction speed adjustment module is specifically used for:

[0036] When the consumption status of hardware resources is busy, determine whether there is a host business;

[0037] When there is no host business, the current round of data reconstruction speed adjustment ends;

[0038] When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

[0039] The method and system for adaptively adjusting data reconstruction speed of the present invention have the following beneficial effects:

[0040] 1. Control the data reconstruction speed for a single storage node in the cluster, refine the flow control granularity to a single disk on a single storage node, make data reconstruction flow control more accurate, and calculate the flow control benefits in real time.

[0041] 2. Perform flow control adjustments on each storage node in the cluster and each disk on each storage node. The flow control granularities do not affect each other, making the flow control of data reconstruction more independent.

[0042] 3. By accurately calculating the hardware resources occupied by each disk on the storage node, independently performing hardware busyness detection and flow control, fully utilizing the storage node hardware resources, and improving the efficiency of data reconstruction without affecting the host business

[0043] 4. Improve the automation and real-time response efficiency of data reconstruction. The entire data reconstruction flow control process is fully automated without the need for operation and maintenance intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 Flow chart of a method for adaptively adjusting data reconstruction speed according to an exemplary first embodiment of the present invention.

[0046] Figure 2 Flow chart of a method for adaptively adjusting data reconstruction speed according to an exemplary second embodiment of the present invention.

[0047] Figure 3 Flow chart of a method for adaptively adjusting data reconstruction speed according to a third exemplary embodiment of the present invention.

[0048] Figure 4 FIG. 4 is an architecture diagram of a system for adaptively adjusting data reconstruction speed according to a fourth exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0049] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0050] It should be noted that the following embodiments and features in the embodiments may be combined with each other in the absence of conflict; and, based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making any creative work are within the scope of protection of the present disclosure.

[0051] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein may be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described herein may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described herein may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein may be used to implement this device and / or practice this method.

[0052] The technical principles of the present invention are as follows:

[0053] The storage cluster automatically collects statistics in real time on the host services and data reconstruction services carried by each storage node in the current cluster, as well as the current consumption of hardware resources by each storage node. By analyzing the busyness of the storage node hardware resources and the changing trends of the host and data reconstruction services carried by the storage nodes, the data reconstruction speed on each storage node is adjusted in real time. When data reconstruction occurs in the cluster, the reconstruction service can use node resources more effectively and improve the efficiency of data reconstruction without affecting the host service; when the storage node is busy, it can automatically adjust the data reconstruction service to release related resources and give priority to the completion of the host service.

[0054] The token bucket model adopted by the present invention is a flow control model, and its use method in data reconstruction is as follows: the tokens in the token bucket are added into the bucket at a set speed; each token corresponds to a certain amount of reconstructed data; before reconstructing a data, the corresponding number of tokens needs to be taken out from the token bucket; if there are not enough tokens, it is necessary to wait for the tokens to be replenished before obtaining the tokens; theoretically, after using the token bucket, the actual data reconstruction speed will be less than or equal to the token bucket speed, and when the actual data reconstruction speed is greater than or equal to the token bucket speed, after using the token bucket, the actual data reconstruction speed is equal to the speed of adding tokens in the token bucket.

[0055] Figure 1 FIG. 1 is a flow chart of a method for adaptively adjusting data reconstruction speed according to a first exemplary embodiment of the present invention. Figure 1 As shown, the method of this embodiment includes:

[0056] Step S1: Obtaining service information of a storage node that needs to be restored and information about hardware resource consumption of the storage node;

[0057] Step S2: judging the consumption state of the hardware resources as busy or idle based on the service information of the storage node that needs to be restored and the consumption information of the hardware resources by the storage node;

[0058] Step S3: adjusting the data reconstruction speed according to the consumption status of hardware resources.

[0059] In step S1 of the method of this embodiment, the service information of the storage node includes: the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk; the consumption information of the hardware resources of the storage node includes: the CPU cores occupied by each disk.

[0060] Figure 2 is a flow chart of a method for adaptively adjusting data reconstruction speed according to an exemplary second embodiment of the present invention. Figure 1 The preferred embodiment of the method shown is as follows Figure 2 As shown, step S2 of the method of this embodiment includes:

[0061] The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy.

[0062] Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing;

[0063] Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, determine that the storage node network is busy;

[0064] When one of the CPU cores occupied by the disk, disk reading and writing, or storage node network is judged to be busy, the consumption status of the hardware resource is judged to be busy. When none of the three items of CPU cores occupied by the disk, disk reading and writing, and storage node network are judged to be busy, the consumption status of the hardware resource is judged to be idle.

[0065] Figure 3 is a flow chart of a method for adaptively adjusting data reconstruction speed according to a third exemplary embodiment of the present invention. Figure 1 A preferred embodiment of the method is shown.

[0066] like Figure 3As shown, step S3 of the method of this embodiment includes:

[0067] When the consumption status of hardware resources is busy, determine whether there is a host business;

[0068] When there is no host business, the current round of data reconstruction speed adjustment ends;

[0069] When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

[0070] Step S3 of the method of this embodiment further includes:

[0071] When the consumption status of hardware resources is idle, the data recovery speed is improved and it is determined whether there is host business;

[0072] When there is no host business, the token bucket model is used to generate the theoretical data reconstruction speed. The actual data reconstruction speed is compared with the theoretical data reconstruction speed. When the two are consistent, the data recovery speed is further improved. When the two are inconsistent, the current round of data reconstruction speed adjustment is terminated.

[0073] When there is a host business, monitor the processing speed of the host business. When the processing speed of the host business does not decrease, generate the theoretical data reconstruction speed through the token bucket model, compare the actual data reconstruction speed with the theoretical data reconstruction speed, and when the two are consistent, continue to improve the data recovery speed. When the two are inconsistent, end this round of data reconstruction speed adjustment; when the processing speed of the host business decreases, end this round of data reconstruction speed adjustment.

[0074] In practical applications, the method of this embodiment controls the data reconstruction speed for a single storage node in the cluster, and refines the flow control granularity to a single disk on a single storage node, making the data reconstruction flow control more accurate and calculating the flow control benefit in real time. Flow control is adjusted for each storage node in the cluster and each disk on each storage node, and each flow control granularity does not affect each other, making the flow control of data reconstruction more independent, that is, the minimum granularity of data reconstruction adjustment is the disk, and the flow control logic between different disks is independent and does not affect each other.

[0075] Figure 4 FIG. 4 is an architecture diagram of a data reconstruction speed adaptive adjustment system according to an exemplary fourth embodiment of the present invention. Figure 4 As shown, the system of this embodiment includes:

[0076] A storage node information acquisition module is used to acquire the service information of the storage node that needs to be restored and the consumption information of the hardware resources of the storage node;

[0077] A hardware resource consumption state judgment module is used to judge the consumption state of the hardware resource as busy or idle based on the service information of the storage node to be restored and the consumption information of the hardware resource by the storage node;

[0078] The data reconstruction speed adjustment module is used to adjust the data reconstruction speed according to the consumption status of hardware resources.

[0079] In practical applications, in the data reconstruction speed adaptive adjustment system of this embodiment, the storage node information acquisition module is specifically used to obtain the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, and obtain the CPU cores occupied by each disk.

[0080] In practical applications, in the adaptive adjustment system of data reconstruction speed in this embodiment, the hardware resource consumption state judgment module is specifically used for:

[0081] The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy.

[0082] Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing;

[0083] Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, determine that the storage node network is busy;

[0084] When one of the CPU cores occupied by the disk, disk reading and writing, or storage node network is judged to be busy, the consumption status of the hardware resource is judged to be busy. When none of the three items of CPU cores occupied by the disk, disk reading and writing, and storage node network are judged to be busy, the consumption status of the hardware resource is judged to be idle.

[0085] In practical applications, in the adaptive adjustment system of data reconstruction speed in this embodiment, the data reconstruction speed adjustment module is specifically used for:

[0086] When the consumption status of hardware resources is busy, determine whether there is a host business;

[0087] When there is no host business, the current round of data reconstruction speed adjustment ends;

[0088] When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

[0089] In practical applications, the method of this embodiment accurately calculates the hardware resources occupied by each disk on the storage node, independently performs hardware busyness detection and flow control, fully utilizes the hardware resources of the storage node, and improves the efficiency of data reconstruction without affecting the host business.

[0090] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

Claims

1. A method for adaptively adjusting data reconstruction speed, characterized in that: The method comprises: Step S1: Obtaining service information of a storage node that needs to be restored and information about hardware resource consumption of the storage node; Step S2: judging the consumption state of the hardware resources as busy or idle based on the service information of the storage node that needs to be restored and the consumption information of the hardware resources by the storage node; Step S3: adjusting the data reconstruction speed according to the consumption status of hardware resources; In step S1, the service information of the storage node includes: the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk; Step S3 further includes: When the consumption status of hardware resources is idle, the data recovery speed is improved and it is determined whether there is host business; When there is no host business, the data reconstruction theoretical speed is generated through the token bucket model, and the actual data reconstruction speed is compared with the data reconstruction theoretical speed. When the two are consistent, the data recovery speed is continued to be improved. When the two are inconsistent, the current round of data reconstruction speed adjustment is ended; wherein, the data reconstruction theoretical speed is the speed of adding tokens to the token bucket; When there is a host service, monitor the processing speed of the host service. When the processing speed of the host service does not decrease, generate the theoretical data reconstruction speed through the token bucket model, compare the actual data reconstruction speed with the theoretical data reconstruction speed, and when the two are consistent, continue to improve the data recovery speed. When the two are inconsistent, end the current round of data reconstruction speed adjustment; when the processing speed of the host service decreases, end the current round of data reconstruction speed adjustment; Step S2, comprising: when one of the CPU cores occupied by the disk, disk read / write, or storage node network is determined to be busy, the consumption state of the hardware resource is determined to be busy; when none of the CPU cores occupied by the disk, disk read / write, and storage node network is determined to be busy, the consumption state of the hardware resource is determined to be idle; Step S3 includes: When the consumption status of hardware resources is busy, determine whether there is a host business; When there is no host business, the current round of data reconstruction speed adjustment ends; When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

2. The method for adaptively adjusting data reconstruction speed according to claim 1, characterized in that: Step S2 comprises: The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy. Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing; Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, and compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with the set threshold. When the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, it is determined that the storage node network is busy.

3. A data reconstruction speed adaptive adjustment system, characterized in that: The system comprises: A storage node information acquisition module is used to acquire the service information of the storage node that needs to be restored and the consumption information of the hardware resources of the storage node. The service information of the storage node includes: the bandwidth of the host service carried by each disk and the bandwidth of the data reconstruction service; A hardware resource consumption state judgment module is used to judge the consumption state of the hardware resource as busy or idle based on the service information of the storage node to be restored and the consumption information of the hardware resource by the storage node; A data reconstruction speed adjustment module, used to adjust the data reconstruction speed according to the consumption status of hardware resources; When the consumption status of hardware resources is idle, the data recovery speed is improved and it is determined whether there is host business; When there is no host business, the data reconstruction theoretical speed is generated through the token bucket model, and the actual data reconstruction speed is compared with the data reconstruction theoretical speed. When the two are consistent, the data recovery speed is continued to be improved. When the two are inconsistent, the current round of data reconstruction speed adjustment is ended; wherein, the data reconstruction theoretical speed is the speed of adding tokens to the token bucket; When there is a host service, monitor the processing speed of the host service. When the processing speed of the host service does not decrease, generate the theoretical data reconstruction speed through the token bucket model, compare the actual data reconstruction speed with the theoretical data reconstruction speed, and when the two are consistent, continue to improve the data recovery speed. When the two are inconsistent, end the current round of data reconstruction speed adjustment; when the processing speed of the host service decreases, end the current round of data reconstruction speed adjustment; The hardware resource consumption state judgment module is specifically used to: when one of the CPU cores occupied by the disk, disk reading and writing, or storage node network is judged to be busy, the hardware resource consumption state is judged to be busy; when the CPU cores occupied by the disk, disk reading and writing, and storage node network are not judged to be busy, the hardware resource consumption state is judged to be idle; The data reconstruction speed adjustment module is specifically used for: When the consumption status of hardware resources is busy, determine whether there is a host business; When there is no host business, the current round of data reconstruction speed adjustment ends; When there is a host business, reduce the data reconstruction speed and monitor the bandwidth of the host business. When the bandwidth of the host business does not increase, end this round of data reconstruction speed adjustment. When the bandwidth of the host business increases, compare the data reconstruction speed with the minimum threshold. When the data reconstruction speed is greater than the minimum threshold, continue to reduce the data reconstruction speed. When the data reconstruction speed is not greater than the minimum threshold, end this round of data reconstruction speed adjustment.

4. The data reconstruction speed adaptive adjustment system according to claim 3, characterized in that: The hardware resource consumption status judgment module is specifically used for: The proportion of the time consumed by each disk in executing the task of the CPU core is calculated according to the consumption information of the hardware resources. When the proportion of the time consumed by the CPU core in executing the task is greater than 80%, it is determined that the CPU core occupied by the disk is busy. Calculate the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk, compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by each disk with a set threshold, and when the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service carried by the disk reaches the set threshold, determine that the disk is busy reading and writing; Summarize the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node, and compare the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node with the set threshold. When the sum of the bandwidth of the host service and the bandwidth of the data reconstruction service of the entire storage node reaches the set threshold, it is determined that the storage node network is busy.

Images