WO2019119833A1 - Storage management method and device, solid state drive and readable storage medium - Google Patents

Abstract

A storage management method and device, a solid state drive and a readable storage medium. The method comprises: when it is detected that data transmitted by a storage channel needs to be stored, acquiring an available storage block and writing the transmitted data into the available storage block (S120); when data writing is completed, detecting whether a storage space state identifier indicates a non-overwritten state (S130); when same indicates the non-overwritten state, detecting whether a storage space reaches a fully-written state (S140); and when the storage space reaches the fully-written state, switching the storage space state identifier to an overwritten state, and starting a timer, and regularly carrying out storage space recovery allocation processing according to the storage life coefficient of each storage block in the storage space and the life balance policy of the storage block (S150). Thus, storage blocks in the storage space can be erased evenly, the overall service life of the solid state drive is improved, and the data read-write efficiency is improved.

Description

Storage management method, device, solid state hard disk and readable storage medium

Cross-reference to related applications

The present application claims priority to Chinese Patent Application No. 201711380674.9, entitled "Storage Management Method, Solid State Drive, and Readable Storage Media", filed on Dec. 20, 2017, the entire contents of In this application.

Technical field

The present application relates to the field of storage technologies, and in particular, to a storage management method, apparatus, solid state hard disk, and readable storage medium.

Background technique

Solid State Drives (SSDs) are hard disks made of solid-state electronic storage chip arrays. They are usually composed of a main control chip and a flash memory chip. The main control chip is the control unit responsible for controlling data storage in the solid state drive. The flash memory chip is The storage unit responsible for storing data in the SSD. Among them, the core of the main control chip is FTL (Flash Translation Layer), FTL is responsible for address mapping, garbage collection, wear leveling, bad block management, etc. The quality of FTL algorithm determines the performance and efficiency of SSD. The pros and cons of life and stability. A flash chip is a long-lived non-volatile memory that retains stored data information in the event of a power outage. Data deletion is in fixed blocks. Unlike a mechanical hard disk, each time a flash chip writes new data, it needs to erase the original data before writing new data.

In order to make SSDs have a longer life cycle, avoid unbalanced erasing (that is, some blocks of SSDs are frequently erased, while others are rarely erased), resulting in SSDs. The overall life cycle is shortened, and technicians in the storage space have proposed wear-leveling algorithms (Wear-Levelling) and garbage collection strategies (Garbage Collection) to solve these problems. The garbage collection strategy is mainly a defragmentation program in the SSD, which can combine the fragmented storage space into a whole block of storage space. The wear leveling algorithm is an algorithm for selecting a memory cell to be erased in a solid state hard disk, so that the erasure times of all memory blocks are as balanced as possible.

However, in the field of data monitoring, since the data acquisition device performing the monitoring task needs to continuously write the collected data to the solid state hard disk, the storage space of the solid state hard disk is always in a nearly full storage state, and it is necessary to utilize the above garbage collection strategy and the above-mentioned wear and tear. The equalization algorithm cyclically rewrites the memory blocks of the solid state drive. Although the standard chip of the traditional standard SSD has a standard garbage collection strategy and wear leveling algorithm, when the standard SSD is shipped from the factory, the FTL architecture of the main control chip has been set and cannot be changed at will. When a standard SSD is used to monitor storage services, the standard FTL architecture cannot meet the actual service storage requirements, and the storage performance is not superior. Moreover, garbage collection and wear leveling will affect the read/write efficiency of SSDs. When the storage space of SSDs is near full, the garbage recovery strategy and wear leveling algorithm with insufficient performance will not improve the storage speed, but will limit the SSD. Storage efficiency.

Application content

The purpose of embodiments of the present application includes, for example, providing a storage management method, apparatus, solid state hard disk, and readable storage medium.

In a first aspect, an embodiment of the present application provides a storage management method, where the method includes:

When it is detected that the data incoming by the storage channel needs to be stored, the available storage block is obtained, and the incoming data is written into the available storage block;

When the data writing is completed, detecting whether the storage space status identifier is a non-overwrite status;

When it is in the non-overwrite state, it is detected whether the storage space reaches the full state;

When the storage space reaches the full state, the storage space state identifier is switched to a replication state, and a timer is started, and the storage space is periodically timed according to the storage life coefficient of each storage block in the storage space and the life balance balancing strategy of the storage block. Recycling distribution processing.

In a second aspect, the embodiment of the present application provides a storage management apparatus, where the apparatus includes:

a data writing module configured to: when detecting that data to be input to the storage channel is to be stored, obtain an available storage block, and write the incoming data into the available storage block;

The first detecting module is configured to detect, when the data writing is completed, whether the storage space status identifier is in a non-overwritten state;

a second detecting module configured to detect whether the storage space reaches a full state when the state is non-overwritten;

Recycling the allocation module, configured to switch the storage space state identifier to a replication state when the storage space reaches a full state, and start a timer according to a storage life coefficient of each storage block in the storage space and a life of the storage block The balancing strategy periodically performs storage space reclamation allocation processing.

In a third aspect, an embodiment of the present application provides a solid state hard disk, including a control unit and a nonvolatile storage unit storing a plurality of computer instructions. When the computer instruction is executed by the control unit, the control unit performs the foregoing Storage management method.

In a fourth aspect, an embodiment of the present application provides a readable storage medium, where the readable storage medium includes a computer program, and the computer program controls the SSD in which the readable storage medium is located to execute the foregoing storage management method.

Compared with the prior art, the present application has the following beneficial effects:

The application provides a storage management method, device, solid state hard disk and readable storage medium. The method includes, when it is detected that data incoming to the storage channel needs to be stored, obtaining an available storage block and writing the incoming data to the available storage block. When the data writing is completed, it is detected whether the storage space status flag is non-overwritten. When it is in the non-overwrite state, it detects whether the storage space has reached the full state. When the storage space reaches the full state, the storage space state identifier is switched to the replication state, and a timer is started, and the storage space collection and distribution process is performed according to the storage life coefficient of each storage block in the storage space and the life balance balancing policy of the storage block. Therefore, the memory blocks in the storage space can be erased evenly, the overall service life of the solid state hard disk is improved, and the data reading and writing efficiency is improved.

DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings to be used in the embodiments will be briefly described below. It should be understood that the following drawings show only certain embodiments of the present application, and therefore It should be seen as a limitation on the scope, and those skilled in the art can obtain other related drawings according to these drawings without any creative work.

FIG. 1 is a block schematic diagram of a monitoring system provided by an embodiment of the present application.

2 is a block schematic diagram of a solid state drive provided by an embodiment of the present application.

FIG. 3 is a flow chart of steps of a storage management method provided by an embodiment of the present application.

FIG. 4 is a second flowchart of the steps of the storage management method provided by the embodiment of the present application.

FIG. 5 is a flowchart of a sub-step of step S120 shown in FIG. 3 according to an embodiment of the present application.

FIG. 6 is one of the flowcharts of the sub-steps of step S150 shown in FIG. 3 provided by the embodiment of the present application.

FIG. 7 is a second flowchart of the sub-steps of step S150 shown in FIG. 3 provided by the embodiment of the present application.

FIG. 8 is a flow chart showing the sub-steps of sub-step S1570 shown in FIG. 6 provided by the embodiment of the present application.

FIG. 9 is a flow chart showing the sub-steps of sub-step S1580 shown in FIG. 6 provided by the embodiment of the present application.

FIG. 10 is a flow chart showing the sub-steps of sub-step S1590 shown in FIG. 6 provided by the embodiment of the present application.

FIG. 11 is a structural block diagram of a storage management apparatus according to an embodiment of the present application.

Icons: 10-monitoring system; 100-storage device; 110-solid state hard disk; 111-control unit; 112-storage unit; 200-data acquisition device.

Detailed ways

The technical solutions in the embodiments of the present application are clearly and completely described in the following with reference to the drawings in the embodiments of the present application. It is obvious that the described embodiments are a part of the embodiments of the present application, and not all of the embodiments. The components of the embodiments of the present application, which are generally described and illustrated in the figures herein, may be arranged and designed in various different configurations. The detailed description of the embodiments of the present application, which is set forth in the claims All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present application without departing from the inventive scope are the scope of the present application.

It should be noted that similar reference numerals and letters indicate similar items in the following figures, and therefore, once an item is defined in a drawing, it is not necessary to further define and explain it in the subsequent drawings.

Please refer to FIG. 1. FIG. 1 is a block diagram of a monitoring system 10 according to an embodiment of the present application. The monitoring system 10 includes at least one data collection device 200 and a storage device 100. The data collection device 200 is in communication with the storage device 100, and the specific communication connection manner may be a wired connection manner or a wireless connection manner. FIG. 1 is a schematic diagram showing a storage device 100 and a plurality of data collection devices 200. Each data collection device 200 can correspond to one storage channel. Each data collection device 200 can transmit data to be stored to the storage device through a corresponding storage channel. 100 to save to the solid state hard disk 110.

In this embodiment, the storage device 100 includes at least one solid state drive 110 configured to store data information. The data collection device 200 performing the monitoring task needs to continuously collect data information, and transmits the collected data information to the storage device 100 through the storage channel to save the data information to the solid state hard disk 110.

In this embodiment, the data collection device 200 may include, but is not limited to, an image acquisition device (eg, a camera, a camera, etc.), an audio collection device (eg, a recorder), and the like.

Please refer to FIG. 2. FIG. 2 is a schematic block diagram of a solid state hard disk 110 according to an embodiment of the present application. The solid state hard disk 110 includes a control unit 111 and a storage unit 112. Wherein, the control unit 111 includes a main control chip configured to control data storage, and the storage unit 112 includes a non-volatile flash memory chip configured to store data information.

For the consumer market, traditional standard SSDs can meet demand due to the simple and uniform storage requirements. For the enterprise market, storage requirements are complex, there are different operating systems and motherboard architectures, and traditional standard SSDs are difficult to meet. Users want to be able to customize FTL and design efficient FTL based on their own data characteristics. For example, the search engine can associate the index table with the physical address of the SSD, and the log data can be directly streamed into the internal flash channel of the SSD, and the database hopes that the key-value can correspond to the physical address of the SSD. Under the premise of this background, CNEXlabs proposed the concept of Open Channel SSD. The control unit of SSD opens the internal channel of SSD to users. The control unit is only responsible for Flash data transmission, ECC (Error Correcting Code, error checking and correction). RAID (Redundant Arrays of Independent Disks), garbage collection, error handling, bad block management, etc., and the design of the FTL layer is implemented by users according to their own needs.

In this embodiment, the solid state drive 110 provided by the solution is preferably implemented by using the Open Channel SSD architecture described above. The advantages of using the Open Channel SSD architecture are: the FTL of the main control chip can be highly integrated with the product driver and file system to improve work efficiency; the storage service defines the FTL according to the requirements, and the FTL can be continuously updated and upgraded according to requirements; different storage can be developed for different applications. Business; can centrally manage multiple SSDs, combine efficient SSD arrays; reduce the load on the SSD master chip, use lower frequency and smaller memory, reduce power consumption and cost.

In this embodiment, the Open Channel SSD architecture can be adopted, combined with the monitoring service requirement, the cyclic rewriting of the monitoring data is directly implemented on the main control chip side, and efficient data garbage collection and wear balance are realized in the process of cyclic overwriting.

In this embodiment, a plurality of computer instructions are stored in the storage unit 112. When the computer instruction is executed by the control unit 111, the control unit 111 may perform the storage management method described below.

The application provides a readable storage medium comprising a computer program. The solid state hard disk 110 in which the readable storage medium is located is controlled to execute the following storage management method when the computer program is running.

The application provides a storage management method. Please refer to FIG. 3. FIG. 3 is a flow chart of the steps of the storage management method provided by the embodiment of the present application. The storage management method is applied to the above-described solid state hard disk 110. The storage management method flow is described in detail below.

Step S120: When it is detected that the data incoming by the storage channel needs to be stored, the available storage block is obtained, and the incoming data is written into the available storage block. If the number of available storage blocks is multiple, the available storage blocks for writing the data may be randomly selected from the plurality of available storage blocks, and of course, the writing for the size of the incoming data may be determined for writing. The number of available storage blocks for this data.

Generally, the storage space of a solid state hard disk is divided into several equally divided storage unit blocks (that is, storage blocks) according to physical addresses. The storage block is a basic unit of data management, and all data to be stored is stored in the storage block. Normally, the storage block has the following states:

1) Idle state: there is no valid data in the memory block, and the data in the memory block has been erased, and the memory block can write new data;

2) Valid status: The valid data (index, video, picture, metadata, log, alarm record, etc.) is stored in the memory block.

3) Invalid state: The data storage duration in the memory block has exceeded the length of time actually needed to be saved, and the memory block can be overwritten, but since the data in the memory block has not been erased, it cannot be used to write new data.

The size of the storage block can be reasonably defined according to actual needs. Since the rewriting speeds of different storage channels are inconsistent, the size of the storage block can be defined according to the storage channel with the slowest replication speed.

Step S130, when the data writing is completed, detecting whether the storage space status identifier is a non-overwrite status.

In the present embodiment, when a data write operation is completed, the control unit 111 needs to detect whether the storage space status flag is a non-overwrite state. The storage space status identifier may be represented by numbers and/or characters, and different numbers and/or characters are used to represent different states. The control unit 111 can determine the current state of the storage space by reading the storage space status identification in the specified location. The control unit 111 knows in advance the storage space state identifier corresponding to the non-overwrite state, and can determine the current storage space state identifier by comparing whether the current storage space state identifier and the storage space state identifier corresponding to the known non-overwrite state are consistent. Whether it is non-overwritten. In the embodiment, the non-overwrite state refers to that when the SSD 110 is initially enabled, there are more available storage blocks in the storage space of the SSD 110, and the initialization state of the storage block is not required to be recycled and rewritten.

Step S140, when it is in the non-overwrite state, it is detected whether the storage space reaches the full state.

In this embodiment, when in the non-overwrite state, the control unit 111 needs to detect whether the storage space reaches the full state.

In this embodiment, the manner in which the control unit 111 detects whether the storage space reaches the full state may include, but is not limited to: 1) detecting whether the number of storage blocks currently occupied by the data in the storage space exceeds the storage full state threshold. If the full state threshold is exceeded, it can be determined that the full state is reached, and if not, the full state is not reached. 2). Detect whether the number of memory blocks currently in the idle state in the storage space is lower than the storage lower threshold. If it is lower than the storage lower threshold, it can be determined that the full state is reached, otherwise the full state is not reached. The storage full state threshold or the storage lower threshold may be set according to actual needs. In the actual application, the above 1) and/or 2) can be flexibly selected, or other detection modes of the full state can be selected, and details are not described herein again.

Step S150, when the storage space reaches the full state, the storage space state identifier is switched to a replication state, and a timer is started, according to a storage life coefficient of each storage block in the storage space and a life balance policy timing of the storage block. Perform storage space reclamation allocation processing.

In this embodiment, the alarm time of the timer can be set so that the storage space collection and distribution process is performed under the trigger of the timer when the alarm time is reached. When setting the alarm time, the alarm duration can be set in advance, and the alarm time of the timer can be determined according to the current time and the alarm duration.

Please refer to FIG. 4. FIG. 4 is a second flowchart of the steps of the storage management method provided by the embodiment of the present application. Before the step S120 of data writing, the method further needs to perform initialization processing on the solid state hard disk 110. The step of performing the initialization process includes: step S110, step S112, step S114, step S116, and step S118.

In step S110, the storage space is divided into a plurality of storage blocks. That is, the storage space is divided into a plurality of storage areas.

In this embodiment, when the SSD 110 initialization process is performed, the storage space of the SSD 110 needs to be divided into a plurality of halved storage blocks according to the physical address.

In this embodiment, the storage block is a basic unit of data management, and all data to be stored needs to be written into a certain storage block for storage. Among them, the memory block mainly includes three states: 1. idle state: data is not saved, and data can be written. 2. Valid status: Save valid data (for example, index data, video data, picture data, metadata, log data, alarm records, etc.), and data cannot be written. 3. Invalid state: Invalid data is saved (that is, the data storage time has exceeded the actual storage time required), and can be overwritten, but since the invalid data is not erased, new data cannot be written yet.

In this embodiment, the monitoring data that the monitoring system 10 needs to store may include, but is not limited to, service data (such as video recording, snapshot pictures, alarm messages, log records, etc.), management data (eg, address mapping, storage block status). , storage channel status, etc.) and index data corresponding to business data. Each of the data collection devices 200 in the monitoring system 10 can correspond to one storage channel, and each data collection device 200 can transmit data to be stored to the storage device 100 through a corresponding storage channel to be saved to the solid state hard disk 110. in.

In this embodiment, when multiple storage channels are saved at the same time, the size of the storage blocks needs to be defined according to specific services, so that each storage channel can flexibly specify the storage space to be occupied according to requirements. For example, if the rewriting speeds of different storage channels are inconsistent, the size of the storage block can be defined according to the storage channel with the slowest replication speed. Therefore, each storage channel can select the number of storage blocks according to the speed of the rewriting speed. For example, a storage channel with a fast rewriting speed can acquire more storage blocks, so as to quickly complete data writing processing and improve storage processing efficiency.

In this embodiment, the rewriting speed of the index data is much slower than the service data. If the size of the storage block is defined according to the rewriting speed of the index data, since the index data occupies less space, the storage block becomes very small, thereby Affects storage efficiency and is not conducive to management. At this time, this embodiment proposes one or more of the following strategies for optimization:

Strategy 1: Index data is not saved separately and can be stored together with business data.

Strategy 2: Combine the indexes of multiple storage channels for storage, increasing the speed of replication.

Strategy 3: Data cycle replication and wear leveling are implemented by means of redundant cache blocks. For example, the B block is used as the redundant buffer block of the A block, and the index data is alternately written into the A and B blocks. Each storage channel has a "number of rewritable memory blocks" parameter, and each time the overwrite is performed, the overwrite operation can be performed according to the number of rewritable memory blocks corresponding to the storage channel. Wherein, in some embodiments, the number of rewritable memory blocks = memory channel copy speed * copy cycle / memory block capacity. Of course, other formulas may also be used to calculate the number of rewritable memory blocks, which are not limited herein. In this embodiment, the saving of management data may be optimized by one or more of the following strategies:

Strategy 1: A dedicated storage area configured to hold management data is separately partitioned at the head of the solid state drive 110, and the dedicated storage area can also be divided into storage blocks of the same size. Since the number of storage blocks is small, the hot and cold exchange processing can be performed with the storage blocks of the remaining storage areas of the solid state hard disk 110 according to a certain period of time, thereby ensuring the overall wear balance.

Strategy 2: Use additional storage media separately (for example, SLC (Single-Level Cell) or MLC (Multi-Level Cell) flash media with higher erasure life and smaller storage space ) Save management data. SLC is characterized by high cost, small capacity and fast speed. MLC is characterized by large capacity, low cost and slow speed.

Step S112, initializing a storage space status identifier, and setting the storage space status identifier to a non-overwrite status.

In this embodiment, an initial storage space state identifier is defined for the storage space of the solid state hard disk 110, and the initialized storage space state identifier is set to a non-overwrite state, that is, an initialization state.

Step S114, initializing one available storage block queue for each storage channel initialization, and adding available storage blocks in each available storage block queue.

In this embodiment, one available storage block queue is configured for each storage channel, and several available storage blocks are added to each available storage block queue. The number of available available storage blocks can be set according to actual needs. Among them, the number of available storage blocks added should not be too much, and it is preferable to add 1-2.

Step S116, adding each storage channel to the list of duplicate channels, and sorting according to the rewriting period from large to small.

In this embodiment, a list of duplicate write channels is created, each storage channel is added to the list of duplicate write channels, and each storage channel is sorted in descending order of the replication cycle from large to small or from small to large.

Step S118, calculating a number of rewritable memory blocks of each memory channel in the list of duplicate write channels, and correspondingly recording the number of rewritable memory blocks into the list of duplicate write channels.

In this embodiment, the formula for calculating the number of rewritable memory blocks may include, but is not limited to, m=s×T÷k. Where m represents the number of rewritable memory blocks, s represents the memory channel copy speed, T represents the overwrite period, and k represents the size of the memory block.

Step S120 and step S150 in Fig. 3 will be described below in conjunction with the above description.

For the step S120, please refer to FIG. 5. FIG. 5 is a flowchart of the sub-steps of step S120 shown in FIG. 3 according to the embodiment of the present application. The step S120 includes a sub-step S121, a sub-step S122, a sub-step S123, a step S124, a sub-step S125, and a sub-step S126.

Sub-step S121, when it is detected that there is an available storage block in the available storage block queue corresponding to the storage channel, obtain a preset number of available storage blocks from the available storage block queue, and write data into the Available in the memory block.

In this embodiment, when detecting that the data incoming in the storage channel needs to be stored, the control unit 111 first detects whether there is an available storage block in the available storage block queue corresponding to the storage channel. When there are available storage blocks, the control unit 111 may obtain a preset number of available storage blocks from the available storage block queue and write data into the available storage blocks. The preset number may be set according to the number of available storage blocks allowed to be stored in the available storage block queue.

In this embodiment, if there is no available storage block in the available storage block queue, the control unit 111 may acquire the storage block in the idle state in the storage space, and use the storage block in the idle state as the available storage block. Join the queue of available storage blocks. If there is no storage block in the storage space in the idle state, it indicates that the current solid state hard disk 110 has no available data storage space, and data writing cannot be performed.

In this embodiment, the available storage block queue does not store an excessive number of available storage blocks, and the solution adopts a strategy of using up and fetching, when the available storage blocks in the available storage block queue are exhausted. Re-adding can effectively avoid the waste of storage resources.

Sub-step S122, when it is detected that the available storage block is full, the available storage block is dequeued from the available storage block queue.

In this embodiment, when one available memory block is full, it is changed to a non-available memory block, and the control unit 111 can move the non-available memory block out of the available memory block queue.

Sub-step S123, it is detected whether the data writing operation is completed. If not, sub-step S124 is performed; if yes, the execution sub-step S121 is returned, that is, when the data writing operation is completed, the control unit 111 cyclically executes the above-described sub-step S121 - sub-step S123.

In this embodiment, considering that the storage capacity of the available storage block is limited, when an available storage block is written, the control unit 111 may detect whether the current data write operation of the storage channel is completed.

Sub-step S124, when the data writing is not completed, the current state of the storage space status identifier is detected. The control unit 111 recognizes whether the current state of the storage space state identifier is the overwrite state or the non-overwrite state when it is determined that the data write is not completed; if it is the non-overwrite state, the sub-step S125 is performed, and if it is the overwrite state, the sub-step S126 is performed.

Sub-step S125, when the current state is a non-overwrite state, acquiring a storage block in an idle state from the storage space to join the available storage block queue, and acquiring a new available storage block from the available storage block queue. Data is written. There is usually a free storage block in the storage space in the non-overwrite state, so that the storage block in the idle state in the storage space can be further utilized, and the free storage block is placed in the available storage block queue to facilitate data writing.

Sub-step S126, when the current state is a replication state, a new available storage block is obtained from the available storage block queue for data writing.

In this embodiment, the available storage blocks are acquired in different manners based on different states of the storage space status identifier.

In this embodiment, when the storage space status is identified as a non-overwrite status, the control unit 111 may acquire a storage block in an idle state from the storage space to join the available storage block queue, and from the available storage. A new available memory block is acquired in the block queue for data writing.

In this embodiment, when the storage space status identifier is a replication status, the control unit 111 periodically performs a storage space collection and allocation process to implement dynamic memory management. The process performing the memory block reclamation allocation can automatically place the reclaimed memory block as an available memory block into the available memory block queue. Thus, the control unit 111 can directly acquire a new available memory block from the available memory block queue for data writing.

In this embodiment, the life balance strategy includes: a hot zone allocation strategy, a cold zone allocation strategy, and a balance zone allocation strategy. The hot zone allocation strategy is a method for recovering and distributing the storage blocks whose storage lifetime coefficient is greater than the hot zone threshold, and the cold zone allocation strategy is a storage lifetime coefficient smaller than that of the present embodiment. The storage block of the area threshold performs the process of collecting and distributing the processing. The balanced area allocation strategy is a method for collecting and distributing the storage block whose storage life coefficient is between the hot zone threshold and the cold zone threshold.

For step S150, in an embodiment, the following steps may be implemented: (1) recording the storage block in the invalid state in the collection list; (2) performing data erasing processing on the storage block recorded in the collection list, Updating the number of erasures of the storage block, updating the status of the storage block to an idle state, and recording to an allocation list; and (3) detecting whether there is an unprocessed pending channel in the to-be-processed channel list; 4) when not present, calculating a storage life coefficient of each storage block according to a current erasure number and an upper erasing number upper limit of each storage block in the allocation list; (5) traversing the allocation list, according to the allocation a storage life coefficient of each storage block in the list, adding each of the storage blocks to a hot zone list, a cold zone list, or a balance zone list; wherein a storage life coefficient of the storage block in the hot zone list is greater than a hot zone a threshold value; a storage life coefficient of the storage block in the cold zone list is smaller than a cold zone threshold; a storage life coefficient of the storage block in the balance zone list is located in the hot zone threshold and the cold zone threshold (6) performing hot zone collection and distribution processing on the storage blocks in the hot zone list according to the hot zone allocation policy; (7) performing cold zone recovery and distribution processing on the storage blocks in the cold zone list according to the cold zone allocation policy; (8) Performing a balance area recovery allocation process on the storage blocks in the balance area list according to the balance area allocation policy.

For ease of understanding, please refer to FIG. 6. FIG. 6 is a flowchart of a sub-step of step S150 shown in FIG. 3 according to an embodiment of the present application, and details a storage life coefficient according to each storage block in the storage space. And the life balance strategy of the storage block periodically performs the specific process of the storage space collection and distribution process. The step S150 includes a sub-step S151, a sub-step S152, a sub-step S153, a sub-step S154, a sub-step S155, a sub-step S156, a sub-step S157, a sub-step S1570, a sub-step S158, a sub-step S1580, a sub-step S159, and a sub-step. Step S1590.

Sub-step S151, the storage channel that is performing the storage service is added to the to-be-processed channel list.

It can be understood that the monitoring system may have multiple data collection devices, and each data collection device corresponds to one storage channel, and the data to be stored may be transmitted to the storage device through the corresponding storage channel. In practical applications, one or more storage channels may be performing storage services during a certain period of time. By monitoring the current state of the storage channel, it is possible to find the storage channel in which the storage service is executed. In this embodiment, the control unit 111 adds each storage channel that is performing the storage service to the to-be-processed channel list.

Sub-step S152: Obtain a channel to be processed from the to-be-processed channel list, and traverse the memory block corresponding to the to-be-processed channel to find a memory block in an active state.

In this embodiment, the control unit 111 may obtain a channel to be processed from the to-be-processed channel list, and then acquire related channel information of the to-be-processed channel, and perform a memory block corresponding to the to-be-processed channel. Traverse, find the memory block that is in a valid state. The control unit 111 may be randomly obtained, for example, a channel to be processed may be randomly obtained from a header, a footer or an arbitrary position of the to-be-processed channel list; of course, the control unit 111 may also be specified. The channel identifier is used to obtain the corresponding pending channel.

Sub-step S153, in the storage block in the active state, acquiring a storage block whose number is the number of the rewritable storage blocks according to a data write time sequence, and setting the acquired state of the storage block to an invalid state, and Record to the recycle list. When the invalid state is set, it may be detected whether the saved duration of the data stored in the obtained storage block exceeds a required storage duration; if yes, it is determined that the acquired data stored in the storage block has expired, and the acquired The status of the memory block is set to an invalid state.

In this embodiment, the control unit 111 may acquire a certain number of storage blocks in the storage block in the active state according to the time sequence of data writing, and the number of the obtained storage blocks is equal to the rewritable storage block. The number of the obtained storage block is set to an invalid state and recorded in the recycle list. That is, the memory block in the invalid state is recorded in the recycle list.

In this embodiment, when the storage block requiring the modified state is acquired in the time sequence of data writing, it is also possible to detect whether the data stored in the acquired storage block has failed, and the data is invalid (that is, the data storage time exceeds the actual The required storage time, for example, when the recording data is saved for longer than the actual recording period to be saved, the state of the storage block is modified and recorded in the recycling list to avoid the storage block in the subsequent recycling process. Valid data erasure.

The above sub-step S151 to sub-step S153 are a specific embodiment for recording the storage block in an invalid state in the collection list.

Sub-step S154, performing data erasing processing on the storage block recorded in the collection list, updating the erasure number of the storage block, updating the status of the storage block to an idle state, and recording to the allocation list.

In this embodiment, each time the memory block is erased, the number of erasures of the memory block needs to be updated, for example, the number of erasures is increased by one. Thereafter, the memory block is switched from the inactive state to the idle state and recorded in the allocation list.

Sub-step S155, detecting whether there is an unprocessed pending channel in the to-be-processed channel list.

In this embodiment, after completing the storage block erasure recovery process of a to-be-processed channel, the control unit 111 may detect whether there is an unprocessed to-be-processed channel in the to-be-processed channel list. If not, execute sub-step S156; if yes, return to execution sub-step S152, that is, when there is still an unprocessed pending channel, the control unit 111 repeatedly performs the above-mentioned sub-step S152, sub-step S153 and sub- Step S154, until all the to-be-processed channels in the to-be-processed channel list are processed.

Sub-step S156, when not present, calculate the storage life coefficient of each storage block according to the current erasure number and the upper limit of the number of erasures of each storage block in the allocation list.

In this embodiment, the calculation method of the storage life coefficient may be, but not limited to, a storage life coefficient=the current erasure number ÷the erasure number upper limit. Each of the storage blocks has an upper limit of the number of erasures, and the upper limit of the number of erasures can be set when the initialization process is performed. The upper limit of the number of erasures of different memory blocks may be the same or different.

In this embodiment, an update manner is adopted in which the memory block is erased once, and the current erasure count is increased by 1. Since the number of erasures affects the service life of the memory block, the upper limit of the erasure number of each memory block is a fixed value. The more the current erasure times, the larger the storage life coefficient of the memory block. Thus, the larger the storage life coefficient indicates that the usable life of the memory block is shorter.

Sub-step S157, traversing the allocation list, adding the storage block whose storage life coefficient is greater than the hot zone threshold to the hot zone list, and sorting the storage blocks in the hot zone list based on the storage life coefficient.

In this embodiment, the control unit 111 traverses the allocation list, adds a storage block whose storage lifetime coefficient is greater than the hot zone threshold to the hot zone list, and sorts the storage blocks according to the storage life coefficient from high to low. Of course, the memory blocks can also be sorted in order of storage life factor from low to high.

Sub-step S1570, performing hot zone collection and distribution processing on the stored blocks in the sorted hot zone list according to the hot zone allocation policy. The storage life coefficient of the storage block in the hot zone list is greater than the hot zone threshold, and the number of erasures is close to the total number of erasures available. The purpose of the hot zone recovery allocation process is to store the block in the hot zone list (which may be simply referred to as a hot zone). Memory block) Quickly update to the balance area.

Sub-step S158, adding the storage block whose storage life coefficient is smaller than the cold zone threshold to the cold zone list, and sorting the storage blocks in the cold zone list based on the storage life coefficient. It can be understood that the hot zone threshold mentioned in sub-step S157 is higher than the cold zone threshold mentioned in sub-step S158.

In this embodiment, the control unit 111 adds a storage block whose storage life coefficient is smaller than the cold zone threshold to the cold zone list, and sorts the storage blocks in order of storage life coefficient from low to high. Of course, the memory blocks can also be sorted in descending order of the storage life factor.

Sub-step S1580, performing cold zone collection and distribution processing on the stored blocks in the sorted cold zone list according to the cold zone allocation policy. The storage life coefficient of the storage block in the cold zone list is smaller than the cold zone threshold, and the number of erasures is much smaller than the total number of erasures available. The purpose of the cold zone recovery allocation process is to store the block in the cold zone list (which may be simply referred to as a cold zone). Memory block) Quickly update to the balance area.

Sub-step S159, adding a storage block whose storage life coefficient is between the hot zone threshold and the cold zone threshold to the balance zone list, and performing, on the storage block in the balance zone list, based on the number of erasures. Sort.

In this embodiment, the control unit 111 stores a storage block with a life coefficient between the hot zone threshold and the cold zone threshold to join the balance zone list, and in the order of the number of erasures from high to low. The storage blocks in the balance area list are sorted. In addition to this, sorting can be performed in descending order of storage life factor.

Sub-step S1590, performing a balance area recovery allocation process on the stored blocks in the sorted balance area list according to the balance area allocation policy. The storage life factor of the storage block in the balance area list is between the hot zone threshold and the cold zone threshold, and the number of erasures is close to the average number of erasures. The purpose of the balance area recovery allocation process is to perform storage block allocation as evenly as possible after the hot zone allocation and the cold zone allocation using the remaining allocatable resources. In this embodiment, in addition to the above steps, the control unit 111 may also periodically count the current number of erasures of each memory block, and calculate the average number of erasures, and the storage life coefficient of each memory block. Statistics are performed and the average storage life factor is calculated.

Sub-step S157 to sub-step S159, that is, a specific method of adding each of the storage blocks to the hot zone list, the cold zone list, or the balance zone list according to the storage life coefficient of each storage block in the allocation list. Implementation.

In this embodiment, the control unit 111 may periodically perform variance calculation based on the current number of erasures and the storage life coefficient, and evaluate the degree of dispersion of the number of erasures and/or the service life based on the variance result. For example, the hot zone threshold and the cold zone threshold can be set based on the degree of dispersion of the variance results. The control unit 111 can define two discrete thresholds (eg, R1, R2, R1 > R2). If the dispersion result is less than R2, the default hot zone threshold and cold zone threshold may be used; if the variance result is more discrete than R1, or the degree of dispersion is within R1 and R2, the degree of dispersion continues to increase. The control unit 111 can dynamically adjust the hot zone threshold and the cold zone threshold to increase the hot zone range and the cold zone range.

Please refer to FIG. 7. FIG. 7 is a second flowchart of the sub-steps of step S150 shown in FIG. 3 according to an embodiment of the present application. Step S150 may further include sub-step S1500 before sub-step S157. The sub-step S1500 only needs to be guaranteed to be executed before the sub-step S157. For example, the sub-step S1500 can also be performed before the sub-step S151.

Sub-step S1500, updating the list of duplicate channels, setting the allocation count of each storage channel to the number of rewritable memory blocks of the storage channel.

In this embodiment, the control unit 111 updates the list of duplicate write channels, traverses the updated list of duplicate write channels, updates the number of rewritable memory blocks of each storage channel in the calculation table, and each storage channel The allocation count is set to the number of rewritable memory blocks of the memory channel, marking the memory channel as pending.

Based on the above description of sub-step S1500, sub-step S1570, sub-step S1580, and sub-step S1590 are described below.

Please refer to FIG. 8. FIG. 8 is a flowchart of a sub-step of the sub-step S1570 shown in FIG. 6 according to the embodiment of the present application, and details a hot zone allocation process, which is an embodiment of a hot zone allocation strategy. The sub-step S1570 includes sub-step S1571, sub-step S1572, sub-step S1573, and step S1574.

Sub-step S1571, obtaining a hot zone storage block from the header of the hot zone list. In this embodiment, since the hot zone list is sorted according to the storage life coefficient from high to low, the control unit 111 sequentially obtains the hot zone storage block from the header of the hot zone list, that is, according to The storage block to be processed is obtained in order of storage life coefficient from large to small.

Sub-step S1572, detecting whether the hot zone storage block has been processed.

In this embodiment, the manner in which the control unit 111 detects whether the hot zone storage block has been processed may include, but is not limited to, detecting whether a valid channel number exists in a current allocation channel number field of the hot zone storage block. If yes, it may be determined that the hot zone storage block has been processed, otherwise it is not processed; or whether the number of the hot zone storage block is recorded in the processed list, and if it is recorded, it may be determined that the hot zone storage block has been Processing, otherwise it is not processed. The valid channel number refers to the channel number of the storage channel.

Sub-step S1573, when not processing, traversing the list of duplicate write channels starting from the header of the list of duplicate write channels, selecting a storage channel whose allocation count is greater than zero, and writing the channel number of the storage channel to the hot zone The current allocation channel number field of the storage block is updated and the allocation count is updated. The storage channel can then also be marked as processed. In practical applications, the way to update the allocation count can be one less than the allocation count.

In this embodiment, since the list of duplicate write channels is sorted according to a rewrite period from large to small, the control unit 111 sequentially selects a storage channel whose allocation count is greater than zero from the header of the list of duplicate write channels, that is, according to The storage channels are obtained in descending order of the replication cycle. Therefore, the control unit 111 can preferentially allocate the storage block with a large storage life coefficient to the storage channel with a long replication period, so that the storage block can be not frequently erased, and the service life of the storage block is extended.

Sub-step S1574, the hot zone storage block is added to the available storage block queue corresponding to the storage channel.

In this embodiment, after completing a hot zone storage block process, the control unit 111 may detect whether there is still an unprocessed hot zone storage block in the hot zone list, and if not, determine the hot zone list. The hot zone storage blocks have been subjected to the hot zone recovery allocation process; if yes, return to sub-step S1571; that is, the control unit 111 repeatedly performs the above sub-step S1571 - sub-step S1574 until the hot zone list The hot zone storage blocks are processed.

Please refer to FIG. 9. FIG. 9 is a flow chart of the sub-steps of the sub-step S1580 shown in FIG. 6 according to the embodiment of the present application, and details a cold zone allocation process, which is an embodiment of the cold zone allocation strategy. The sub-step S1580 includes sub-step S1581, sub-step S1582, sub-step S1583, and step S1584.

Sub-step S1581, obtaining a cold zone storage block from the header of the cold zone list.

In this embodiment, since the cold zone list is sorted according to the storage life coefficient from low to high, the control unit 111 sequentially acquires the cold zone storage block from the header of the cold zone list, that is, according to The storage blocks whose storage life coefficients are small to large are acquired.

Sub-step S1582, detecting whether the cold zone storage block has been processed.

In this embodiment, the manner in which the control unit 111 detects whether the cold zone storage block has been processed is similar to the manner in which the hot zone storage block is detected, and details are not described herein.

Sub-step S1583, when unprocessed, traversing the list of duplicate write channels starting from the end of the list of duplicate write channels, selecting a storage channel whose allocation count is greater than zero, and writing the channel number of the storage channel to the cold zone The currently allocated channel number field of the memory block is updated, and the allocation count is updated (such as decrementing the allocation count by one), and the cold zone storage block is marked as processed.

In this embodiment, the control unit 111 selects a storage channel whose allocation count is greater than zero from the end of the list of the duplicate write channels, that is, acquires the storage channels in descending order of the rewrite period. Therefore, the control unit 111 can preferentially allocate a storage block with a small storage life coefficient to a storage channel with a short replication cycle, that is, preferentially allocate a storage block with a long service life to a storage channel that frequently performs overwriting.

Sub-step S1584, adding the cold zone storage block to the available storage block queue corresponding to the storage channel.

In this embodiment, after completing a cold zone storage block process, the control unit 111 may detect whether there is still an unprocessed cold zone storage block in the cold zone list, and if so, the control unit 111 repeatedly executes. Sub-step S1581 - sub-step S1584, until the cold zone storage blocks in the cold zone list are all processed.

Referring to FIG. 10, FIG. 10 is a flowchart of a sub-step of the sub-step S1590 shown in FIG. 6 according to an embodiment of the present application, and details a balancing area allocation process, which is an embodiment of a balanced area allocation strategy. The sub-step S1590 includes sub-step S1591, sub-step S1592, sub-step S1593, and step S1594.

Sub-step S1591, obtaining a storage channel whose allocation count is not zero from the header of the duplicate channel list. For example, a memory channel whose allocation count is not zero can be obtained as a pending channel from the header order of the copy channel list.

Sub-step S1592, traversing the balance area list from the header of the balance area list, and sequentially acquiring a balance area storage block having a number equal to the allocation count of the storage channel.

Sub-step S1593, writing the channel number of the storage channel into the acquired current allocation channel number of the balance area storage block, and marking the balance area storage block as processed, and counting the allocation of the storage channel Zero.

Sub-step S1594, adding the balance area storage block to the available storage block queue corresponding to the storage channel.

If there is a storage channel in the duplicate channel list whose allocation count is not zero, the execution returns to step S1591, otherwise the flow ends.

In this embodiment, the allocation of the balance area storage block does not need to be finely distributed according to the storage life coefficient like the hot area storage block and the cold area storage block, and the balance area storage block can be batch-distributed according to the storage channel requirement. .

In this embodiment, the scheme adopts the Open Channel SSD architecture, and the user can design the FTL of the solid state hard disk 110 by itself, so that the garbage collection algorithm and the wear leveling algorithm of the solid state hard disk 110 meet their actual needs. According to the storage life factor, the storage block of the solid state hard disk 110 is divided into a hot zone storage block, a cold zone storage block and a balance zone storage block, and the storage blocks of different zones have a corresponding recycling allocation strategy. It can effectively ensure the equalization of the erasure times of all the storage blocks in the storage space, improve the overall service life of the solid state hard disk 110, and improve the data reading and writing efficiency.

Further, the embodiment provides a storage management device, which may be disposed on the SSD side. Referring to one of the structural block diagrams of the storage management device shown in FIG. 11, the device includes:

The data writing module 1101 is configured to: when it is detected that data to be input to the storage channel is to be stored, obtain an available storage block, and write the incoming data into the available storage block;

The first detecting module 1102 is configured to detect, when the data writing is completed, whether the storage space status identifier is in a non-overwritten state;

The second detecting module 1103 is configured to detect whether the storage space reaches a full state when in a non-overwrite state;

The recycling distribution module 1104 is configured to switch the storage space state identifier to a replication state when the storage space reaches a full state, and start a timer according to a storage life coefficient of each storage block in the storage space and a storage block. The life balance strategy periodically performs storage space reclamation allocation processing.

In some embodiments, the apparatus may further include the following modules:

a storage block dividing module configured to divide the storage space into a plurality of storage blocks;

a state setting module configured to initialize an initialization of the storage space state identifier, and set the storage space state identifier to a non-overwrite state;

The storage block adds a module, configured to initialize one storage block queue for each storage channel initialization, and add available storage blocks in each available storage block queue;

The channel sorting module is configured to add each storage channel to the list of duplicate channels, and sort according to a rewriting period from large to small;

The number recording module is configured to calculate a number of rewritable memory blocks of each of the memory channels in the list of duplicate write channels, and record the number of the rewritable memory blocks correspondingly into the list of duplicate write channels.

In a specific implementation, the data writing module is configured to:

When a storage block is available in the available storage block queue corresponding to the storage channel, a preset number of available storage blocks are obtained from the available storage block queue, and data is written into the available storage block. ;

Determining the available memory block from the available memory block queue when it is detected that the available memory block is full;

Check if the data write operation is completed;

When the data writing is not completed, detecting the current state of the storage space status identifier;

When the current state is a non-overwrite state, the storage block that is in the idle state is acquired from the storage space to join the available storage block queue, and a new available storage block is obtained from the available storage block queue for data writing;

When the current state is a replication state, a new available memory block is obtained from the available memory block queue for data writing.

In some embodiments, the life balance strategy includes: a hot zone allocation strategy, a cold zone allocation policy, and a balance zone allocation policy, the recycling allocation module configured to:

Add the storage channel that is performing the storage service to the pending channel list.

Obtaining a to-be-processed channel from the to-be-processed channel list, traversing the storage block corresponding to the to-be-processed channel, and searching for a storage block in an active state;

And storing, in the storage block in an active state, a storage block whose number is the number of the rewritable storage blocks according to a data write time sequence, setting the acquired state of the storage block to an invalid state, and recording to the collection list. ;

Performing data erasing processing on the storage blocks recorded in the collection list, updating the erasure times of the storage blocks, updating the status of the storage blocks to an idle state, and recording to the allocation list;

Detecting whether there is an unprocessed pending channel in the to-be-processed channel list;

When not present, calculating a storage life coefficient of each storage block according to a current erasure number and an upper erasing number upper limit of each storage block in the allocation list;

Traversing the allocation list, adding the storage block whose storage lifetime coefficient is greater than the hot zone threshold to the hot zone list, and sorting the storage blocks in the hot zone list based on the storage life coefficient;

Performing hot zone collection and distribution processing on the storage blocks in the sorted hot zone list according to the hot zone allocation policy;

Adding the storage block whose storage life coefficient is smaller than the cold zone threshold to the cold zone list, and sorting the storage blocks in the cold zone list based on the storage life coefficient;

Performing cold zone collection and distribution processing on the storage blocks in the sorted cold zone list according to the cold zone allocation policy;

And storing a storage block whose storage life coefficient is between the hot zone threshold and the cold zone threshold into a balance zone list, and sorting the storage blocks in the balance zone list based on the number of erasures;

The balanced area recovery allocation process is performed on the stored blocks in the sorted balance area list according to the balance area allocation policy.

The implementation principle and the technical effects of the device provided in this embodiment are the same as those in the foregoing embodiment. For the sake of brief description, where the device embodiment is not mentioned, reference may be made to the corresponding content in the foregoing method embodiment.

In summary, the embodiments of the present application provide a storage management method, an apparatus, a solid state hard disk, and a readable storage medium. The method includes, when it is detected that data incoming to the storage channel needs to be stored, obtaining an available storage block and writing the incoming data to the available storage block. When the data writing is completed, it is detected whether the storage space status flag is non-overwritten. When it is in the non-overwrite state, it detects whether the storage space has reached the full state. When the storage space reaches the full state, the storage space state identifier is switched to a replication state, and a timer is started, and the storage space is periodically timed according to the storage life coefficient of each storage block in the storage space and the life balance balancing strategy of the storage block. Recycling distribution processing. Thus, by dividing the storage blocks of the solid state hard disk into hot zone storage blocks, cold zone storage blocks, and balance zone storage blocks according to the storage life, the storage blocks of different zones have corresponding recycling allocation policies. It can balance the number of erasures of memory blocks in the storage space, improve the overall life of the solid state drive, and improve data read and write efficiency.

The above description is only the preferred embodiment of the present application, and is not intended to limit the present application, and various changes and modifications may be made to the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of this application are intended to be included within the scope of the present application.

Industrial applicability:

By applying the technical solution of the present application, the storage blocks in the storage space can be uniformly erased, the overall service life of the solid state hard disk is improved, and the data reading and writing efficiency is improved.

Claims (19)

1. A storage management method, the method comprising:

   When it is detected that the data incoming by the storage channel needs to be stored, the available storage block is obtained, and the incoming data is written into the available storage block;

   When the data writing is completed, detecting whether the storage space status identifier is a non-overwrite status;

   When it is in the non-overwrite state, it is detected whether the storage space reaches the full state;

   When the storage space reaches the full state, the storage space state identifier is switched to a replication state, and a timer is started, and the storage space is periodically timed according to the storage life coefficient of each storage block in the storage space and the life balance balancing strategy of the storage block. Recycling distribution processing.
2. The method of claim 1, wherein before the obtaining the available memory block and writing the incoming data to the available memory block, the method further comprises:

   Divide storage space into multiple storage blocks;

   Initializing a storage space status identifier, and setting the storage space status identifier to a non-overwrite status;

   Configure an available block queue for each storage channel initialization and add available blocks to each available block queue;

   Add each storage channel to the list of replication channels and sort them from large to small according to the replication cycle;

   Calculating a number of rewritable memory blocks of each memory channel in the list of duplicate write channels, and recording the number of rewritable memory blocks correspondingly into the list of duplicate write channels.
3. The method of claim 2, wherein the obtaining an available storage block and writing the incoming data to the available storage block comprises:

   When a storage block is available in the available storage block queue corresponding to the storage channel, a preset number of available storage blocks are obtained from the available storage block queue, and data is written into the available storage block. ;

   Determining the available memory block from the available memory block queue when it is detected that the available memory block is full;

   Check if the data write operation is completed;

   When the data writing is not completed, detecting the current state of the storage space status identifier;

   When the current state is a non-overwrite state, the storage block that is in the idle state is acquired from the storage space to join the available storage block queue, and a new available storage block is obtained from the available storage block queue for data writing;

   When the current state is a replication state, a new available memory block is obtained from the available memory block queue for data writing.
4. The method according to claim 1, wherein the detecting whether the storage space reaches a full state comprises:

   Detecting whether the number of memory blocks currently occupied by the data in the storage space exceeds a storage full state threshold; if yes, determining that the storage space is full;

   Alternatively, it is detected whether the number of memory blocks currently in the idle state in the storage space is lower than the storage lower threshold; if so, it is determined that the storage space reaches the full state.
5. The method according to any one of claims 2 or 3, wherein the life balance strategy comprises: a hot zone allocation strategy, a cold zone allocation strategy, and a balance zone allocation policy, according to each storage block in the storage space. The storage life factor and the life balance strategy of the storage block are periodically scheduled to perform storage space collection and distribution processing, including:

   Recording the invalid storage block in the recycle list;

   Performing data erasing processing on the storage blocks recorded in the collection list, updating the erasure times of the storage blocks, updating the status of the storage blocks to an idle state, and recording to the allocation list;

   Check whether there are unprocessed pending channels in the pending channel list.

   When not present, calculating a storage life coefficient of each storage block according to a current erasure number and an upper erasing number upper limit of each storage block in the allocation list;

   Traversing the allocation list, adding each of the storage blocks to a hot zone list, a cold zone list, or a balance zone list according to a storage life coefficient of each storage block in the allocation list; wherein, in the hot zone list The storage life coefficient of the storage block is greater than the hot zone threshold; the storage life coefficient of the storage block in the cold zone list is smaller than the cold zone threshold; the storage lifetime coefficient of the storage block in the balance zone list is located in the hot zone threshold and Between the cold zone thresholds;

   And performing hot zone collection and distribution processing on the storage blocks in the hot zone list according to the hot zone allocation policy; performing cold zone collection and distribution processing on the storage blocks in the cold zone list according to the cold zone allocation policy;

   The balance area recovery allocation process is performed on the storage blocks in the balance area list according to the balance area allocation policy.
6. The method according to claim 5, wherein the recording the invalid state storage block in the collection list comprises:

   Add the storage channel that is performing the storage service to the pending channel list.

   Obtaining a to-be-processed channel from the to-be-processed channel list, traversing the storage block corresponding to the to-be-processed channel, and searching for a storage block in an active state;

   And storing, in the storage block in an active state, a storage block whose number is the number of the rewritable storage blocks according to a data write time sequence, setting the acquired state of the storage block to an invalid state, and recording to the collection list. .
7. The method according to claim 5, wherein said adding said said storage block to a hot zone list, a cold zone list or a balance zone list according to a storage life coefficient of each of said storage blocks in said allocation list ,include:

   Adding the storage block whose storage life coefficient is greater than the hot zone threshold to the hot zone list, and sorting the storage blocks in the hot zone list based on the storage life coefficient;

   Adding the storage block whose storage life coefficient is smaller than the cold zone threshold to the cold zone list, and sorting the storage blocks in the cold zone list based on the storage life coefficient;

   And storing a storage block whose storage life coefficient is between the hot zone threshold and the cold zone threshold into a balance zone list, and sorting the storage blocks in the balance zone list based on the number of erasures.
8. The method according to claim 6, wherein the setting the acquired state of the storage block to an invalid state comprises:

   Detecting whether the saved duration of the data stored in the obtained storage block exceeds a required storage time;

   If so, it is determined that the data stored in the acquired storage block has expired, and the acquired state of the storage block is set to an invalid state.
9. The method according to claim 5, wherein before the traversing the allocation list, the storage space recovery allocation processing is performed according to a storage life coefficient of each storage block in the storage space and a life balance policy timing of the storage block. ,Also includes:

   The list of duplicate channel is updated, and the allocation count of each storage channel is set to the number of rewritable memory blocks of the storage channel.
10. The method according to claim 9, wherein the performing hot zone collection and distribution processing on the storage blocks in the hot zone list according to the hot zone allocation policy comprises:

    Obtaining a hot zone storage block from a header of the hot zone list;

    Detecting whether the hot zone storage block has been processed;

    When not processed, traverse the list of duplicate write channels from the header of the list of duplicate write channels, select a storage channel whose allocation count is greater than zero, and write the channel number of the storage channel to the current storage block of the hot zone Assign a channel number field and update the allocation count;

    Adding the hot zone storage block to the available storage block queue corresponding to the storage channel.
11. The method according to claim 10, wherein the detecting whether the hot zone storage block has been processed comprises:

    Detecting whether a valid channel number exists in a current allocation channel number field of the hot zone storage block, and if so, determining that the hot zone storage block has been processed;

    or,

    It is detected whether the number of the hot zone storage block is recorded in the processed list, and if so, it is determined that the hot zone storage block has been processed.
12. The method according to claim 9, wherein the performing cold zone collection and distribution processing on the storage blocks in the cold zone list according to the cold zone allocation policy comprises:

    Obtaining a cold zone storage block from a header of the cold zone list;

    Detecting whether the cold zone storage block has been processed;

    When not processed, traversing the list of duplicate write channels from the end of the list of duplicate write channels, selecting a storage channel whose allocation count is greater than zero, and writing the channel number of the storage channel to the current storage block of the cold storage block Allocating the channel number field, and updating the allocation count, marking the cold zone storage block as processed;

    Adding the cold zone storage block to the available storage block queue corresponding to the storage channel.
13. The method according to claim 9, wherein the balancing area allocation processing is performed on the storage blocks in the balance area list according to the balance area allocation policy, including:

    Obtaining, from a header of the list of duplicate write channels, a storage channel whose allocation count is not zero;

    Traversing the balance area list from the header of the balance area list, and sequentially acquiring a balance area storage block equal to the allocation count of the storage channel;

    Writing a channel number of the storage channel into the acquired current allocation channel number of the balance area storage block, and marking the balance area storage block as processed, and setting the allocation count of the storage channel to zero;

    Adding the balance area storage block to the available storage block queue corresponding to the storage channel.
14. A storage management device, characterized in that the device comprises:

    a data writing module configured to: when detecting that data to be input to the storage channel is to be stored, obtain an available storage block, and write the incoming data into the available storage block;

    The first detecting module is configured to detect, when the data writing is completed, whether the storage space status identifier is in a non-overwritten state;

    a second detecting module configured to detect whether the storage space reaches a full state when the state is non-overwritten;

    Recycling the allocation module, configured to switch the storage space state identifier to a replication state when the storage space reaches a full state, and start a timer according to a storage life coefficient of each storage block in the storage space and a life of the storage block The balancing strategy periodically performs storage space reclamation allocation processing.
15. The device according to claim 14, wherein the device further comprises:

    a storage block dividing module configured to divide the storage space into a plurality of storage blocks;

    a state setting module configured to initialize an initialization of the storage space state identifier, and set the storage space state identifier to a non-overwrite state;

    The storage block adds a module, configured to initialize one storage block queue for each storage channel initialization, and add available storage blocks in each available storage block queue;

    The channel sorting module is configured to add each storage channel to the list of duplicate channels, and sort according to a rewriting period from large to small;

    The number recording module is configured to calculate a number of rewritable memory blocks of each of the memory channels in the list of duplicate write channels, and record the number of the rewritable memory blocks correspondingly into the list of duplicate write channels.
16. The apparatus according to claim 15, wherein said data writing module is configured to:

    When a storage block is available in the available storage block queue corresponding to the storage channel, a preset number of available storage blocks are obtained from the available storage block queue, and data is written into the available storage block. ;

    Determining the available memory block from the available memory block queue when it is detected that the available memory block is full;

    Check if the data write operation is completed;

    When the data writing is not completed, detecting the current state of the storage space status identifier;

    When the current state is a non-overwrite state, the storage block that is in the idle state is acquired from the storage space to join the available storage block queue, and a new available storage block is obtained from the available storage block queue for data writing;

    When the current state is a replication state, a new available memory block is obtained from the available memory block queue for data writing.
17. The device according to any one of claims 15 or 16, wherein the life balance strategy comprises: a hot zone allocation strategy, a cold zone allocation strategy, and a balance zone allocation policy, wherein the recycling allocation module is configured to:

    Add the storage channel that is performing the storage service to the pending channel list.

    Obtaining a to-be-processed channel from the to-be-processed channel list, traversing the storage block corresponding to the to-be-processed channel, and searching for a storage block in an active state;

    And storing, in the storage block in an active state, a storage block whose number is the number of the rewritable storage blocks according to a data write time sequence, setting the acquired state of the storage block to an invalid state, and recording to the collection list. ;

    Performing data erasing processing on the storage blocks recorded in the collection list, updating the erasure times of the storage blocks, updating the status of the storage blocks to an idle state, and recording to the allocation list;

    Detecting whether there is an unprocessed pending channel in the to-be-processed channel list;

    When not present, calculating a storage life coefficient of each storage block according to a current erasure number and an upper erasing number upper limit of each storage block in the allocation list;

    Traversing the allocation list, adding the storage block whose storage lifetime coefficient is greater than the hot zone threshold to the hot zone list, and sorting the storage blocks in the hot zone list based on the storage life coefficient;

    Performing hot zone collection and distribution processing on the storage blocks in the sorted hot zone list according to the hot zone allocation policy;

    Adding the storage block whose storage life coefficient is smaller than the cold zone threshold to the cold zone list, and sorting the storage blocks in the cold zone list based on the storage life coefficient;

    Performing cold zone collection and distribution processing on the storage blocks in the sorted cold zone list according to the cold zone allocation policy;

    And storing a storage block whose storage life coefficient is between the hot zone threshold and the cold zone threshold into a balance zone list, and sorting the storage blocks in the balance zone list based on the number of erasures;

    The balanced area recovery allocation process is performed on the stored blocks in the sorted balance area list according to the balance area allocation policy.
18. A solid state hard disk comprising a control unit and a non-volatile storage unit storing a plurality of computer instructions, wherein the control unit performs any of claims 1-13 when the computer instruction is executed by the control unit A storage management method as described.
19. A readable storage medium, the readable storage medium comprising a computer program, characterized by:

    The solid state hard disk in which the readable storage medium is located is executed to execute the storage management method according to any one of claims 1-13.

Images