TW201423567A - Data writing method and system - Google Patents

Abstract

A data writing method and a system using the same are provided for writing data from a host system into a storage device having a first storage unit and a second storage unit. The method includes providing an application characteristic database. The method also includes receiving a write command from an application running on the host system. The method further includes obtaining a data access pattern corresponding to the application from the application characteristic database, and selecting one of the first storage unit and the second storage unit for writing data therein according to the obtained data access pattern and a management strategy.

Description

Data writing method and system

The present invention relates to a method of writing data, and more particularly to a method and system for writing data to a storage device having a plurality of types of storage units.

The so-called two-disc storage system is a combination of a traditional hard disk drive (HDD) and a solid state disk (SSD). It combines the high capacity of a traditional hard disk with the high speed of a solid state hard disk. Features are gradually being widely used in computer systems.

In detail, although the solid-state hard disk has the advantage of fast reading and writing speed, since the price is higher than that of the conventional hard disk, if only the solid-state hard disk is used as the storage medium, the cost of the storage system is likely to be too high. Therefore, in a storage system that emphasizes high storage capacity, a conventional hard disk is still used as a main storage medium, and a relatively small-capacity solid-state hard disk is used as a data cache. In other words, most of the two-disc storage systems will store the data in a solid-state hard disk with a fast access speed when receiving the data to be written by the computer system. The disc is moved to a traditional hard drive. This ensures that the data access speed and storage capacity are consistent with the needs of the user.

However, because the SSD in the dual-disc storage system can only provide a small storage capacity, how to write the data more effectively so that the user can more clearly feel the advantages of the solid-state hard disk is the relevant industry. Developing and producing One of the important topics in product time.

In view of this, the present invention provides a data writing method and system, which can effectively improve the performance, power consumption, and device life of a storage device having several types of storage units.

The invention provides a data writing method for writing data from a host system to a storage device having a first storage unit and a second storage unit, the method comprising providing an application property database. A write command is received from an application executing on the host system to write data to the storage device. Get the data access type corresponding to the application from the application property database. And selecting data to be written to the first storage unit or the second storage unit according to the data access type and the access management policy.

In an embodiment of the present invention, the access management policy includes a first type of access rule corresponding to the first storage unit and a second type of access rule corresponding to the second storage unit.

In an embodiment of the present invention, the data access type includes an access frequency, a file size, and a single instruction access data amount, and the data is written according to the data access type and the access management policy. The step of entering the first storage unit or the second storage unit includes writing the data to the first storage unit when the access frequency, the file size, and the individual statistical value of the single instruction access data amount conform to the first type of access rule. When the access frequency, the file size, and the individual statistical value of the single instruction access data amount conform to the second type of access rule, the data is written into the second storage unit.

In an embodiment of the invention, the step of providing an application property database includes collecting an application access frequency, a file size, and a plurality of actual values of a single instruction access data amount during a specific period. According to the above actual value, the statistical value of the access frequency, the file size and the single instruction access data amount are calculated. The statistics of the access frequency, the file size, and the single instruction access data amount are recorded in the application property database.

In an embodiment of the invention, the data writing method further comprises updating the data access type corresponding to the application recorded in the application property database according to the content of the write command. Moreover, the first type of access rules and/or the second type of access rules are periodically modified according to the operational performance of the storage device.

In an embodiment of the invention, the first storage unit is a flash memory storage unit, and the first type of access rule includes at least a high access frequency threshold and a random access mode, and the second storage unit is magnetic. A disc storage unit, and the second type of access rule includes at least a large file capacity lower limit value and a continuous access mode.

From another point of view, the present invention provides a data writing system including an application property database and a decision module coupled to each other. The decision module obtains the data access type corresponding to the application from the application property database when receiving the write command executed by the application executed by the host system to write a data to the storage device, and saves the data according to the data. The fetching mode and the access management policy select to write the data to the first storage unit or the second storage unit of the storage device.

In an embodiment of the present invention, the access management policy includes a first type of access rule corresponding to the first storage unit and a second storage unit The second type of access rule should be.

In an embodiment of the invention, the data access type includes an access frequency, a file size, and a single instruction access data amount, and the decision module accesses the frequency, the file size, and the single instruction access. When the individual statistics of the data volume meet the first type of access rules, the data is selected to be written into the first storage unit. Moreover, the decision module selects to write the data to the second storage unit when the access frequency, the file size, and the individual statistic value of the single instruction access data amount meet the second type of access rule.

In an embodiment of the present invention, the data writing system further includes a database building module coupled to the application property database. The database building module collects the actual frequency of the application access frequency, the file size and the single instruction access data amount in a specific period, and calculates the access frequency, the file size and the single instruction access data according to the actual value. The individual statistical values, as well as the individual statistics of the access frequency, file size and single instruction access data, are recorded in the application property database.

In an embodiment of the present invention, the decision module updates the data access type corresponding to the application recorded in the application property database according to the content of the write command, and periodically corrects the first according to the operation performance of the storage device. Class access rules and/or second class access rules.

In an embodiment of the invention, the first storage unit is a flash memory storage unit, and the first type of access rule includes at least a high access frequency threshold and a random access mode, and the second storage unit is magnetic. A disc storage unit, and the second type of access rule includes at least a large file capacity lower limit value and a continuous access mode.

Based on the above, the data writing method and system of the present invention can appropriately select the data to be written into the first storage unit or the second storage unit of the storage device, thereby utilizing the advantages of different types of storage units to enhance the storage device. The overall performance, while achieving the purpose of effectively controlling the power consumption of the storage device and extending the life of the device.

The above described features and advantages of the present invention will be more apparent from the following description.

1 is a schematic diagram of the use of a data writing system according to an embodiment of the invention. Referring to FIG. 1, the data writing system 100 is configured in the host system 10. In this embodiment, the host system 10 may be a desktop computer, a notebook computer, a tablet computer, or various electronic devices having arithmetic processing capabilities, and the types thereof are not limited herein. The operating system 13 is used to manage the hardware and software resources of the host system 10, and the remaining applications 15 ₁ to 15 _n (n is a positive integer) executable in the host system 10 must be used by the operating system 13. The hardware resources of the host system 10.

The storage device 20 can be built into the host system 10 or externally connected to the host system 10 via a transmission line. In detail, the storage device 20 has a first storage unit 21 and a second storage unit 23, wherein the first storage unit 21 and the second storage unit 23 respectively use different types of storage media. For convenience of description, it is assumed in the following embodiments that the first storage unit 21 is a flash memory storage unit using a flash memory as a storage medium, such as a Solid State Disk (SSD). Second storage unit 23 is a disk storage unit using a disk as a storage medium, such as a Hard Disk Drive (HDD). That is to say, the data access speed of the first storage unit 21 is higher than that of the second storage unit 23. However, based on cost considerations, the capacity of the second storage unit 23 in the storage device 20 is greater than that of the first storage unit 21.

The data writing system 100 is used to write data from the host system 10 to the storage device 20, and in particular to write the data generated by the applications 15 ₁ to 15 _n to the storage device 20. As shown in FIG. 1, the data writing system 100 includes an application property database 110, a decision module 120, and a database creation module 130. The decision module 120 and the database creation module 130 are individually coupled to the application property database 110. The functions of the components of the data writing system 100 are described as follows: The application property database 110 is used to record various applications 15 ₁ to 15 _n installed in the host system 10 and their data access types. In the present embodiment, the application property database 110 is created by the database creation module 130 collecting the actual execution status of the applications 15 ₁ to 15 _n . In other embodiments, however, the application property database 110 may also be provided by the manufacturer of the host system 10 and written by the user to the system 100.

The decision module 120 is configured to predict the read and write requirements of the application for the storage device 20 according to the information provided by the application property database 110 when any application executed by the host system 10 is to be written into the storage device 20. And accordingly, it is decided to write the data into the first storage unit 21 or the second storage unit 23. Through the operation of the decision module 120, different types can be used. The advantages of the storage unit. For example, the high access speed of the first storage unit 21 is used to reduce the time for the user to wait for data access, and the storage capacity of the storage device 20 is increased to meet the use of the second storage unit 23. The need for storage capacity.

In this embodiment, the decision module 120 can be a hardware component with processing capability, such as a central processing unit (CPU) or a microprocessor (Microprocessor), or a logic gate and related circuits. . In yet another embodiment, the decision module 120 can also be a software component (eg, an application) having logical operations and processing capabilities. In other embodiments, the decision module 120 can also be a combination of the aforementioned hardware components and software components.

The detailed operation of the data writing system 100 of the present embodiment will be described below with reference to FIG. Please also refer to Figure 1 and Figure 2.

First, as shown in step S210, an application property database 110 is provided. Specifically, since different types of applications often use different data access types, for example, multimedia playback applications access large and continuous files, while document processing applications are relatively large. Frequent access to smaller files. For example, Figure 3 shows the performance of the spreadsheet application, the word processing application, the email application, the newsletter application, and the image editing application for different data access types. Referring to FIG. 3, in such a data access type in which the amount of data read randomly does not exceed 16 kilobytes (KiloByte, KB), the proportion of the word processing application is higher than that of other applications, indicating that the file is processed by the word processing. The data created by the application will be read as a small random read in the future. In the continuous access of such data access types, the proportion of image editing applications is much higher than other applications, indicating that the data created by the image editing application is continuously read in the future. In order to select suitable storage units for the characteristics of each application, for the applications 15 ₁ to 15 _n executable in the host system 10, the database creation module 130 will collect the execution time of each application during a specific period of time. Data access type. For example, the actual value of various data access types such as the access frequency, the file size, and the amount of access data corresponding to a single read or write command are collected. Calculate the statistics of each application's access frequency, file size and single instruction access data according to the actual value, and record the above statistics in the application property database as the application data access. Type.

Then, when an application (for example, the application program 15 ₁ ) executing in the host system 10 wants to write data to the storage device 20 , the operating system 13 transmits the write command and the data to be written to the decision module. 120. Therefore, as shown in step S220, the decision module 120 receives a write command from the application program 15 ₁ to write data to the storage device.

Next, at step S230, the decision module 120 from the application database 110 to obtain characteristic application ₁₅₁ corresponding to the data access type, including an access frequency, a single instruction, file size and access data amount of individual statistics. And as shown in step S240, the decision module 120 selects to write the data to the first storage unit 21 or the second storage unit 23 of the storage device 20 according to the data access type and an access management policy.

In this embodiment, the access management policy includes a first type of access rule corresponding to the first storage unit 21 and a corresponding one of the second storage unit 23 The second type of access rules. Further, when the first storage unit 21 is a flash memory storage unit, the first type of access rule includes at least a high access frequency threshold and a random access mode, and when the second storage unit 23 is a disk storage unit. The second type of access rule includes at least a large file capacity lower limit value and a continuous access mode.

4 is a flow chart of selecting data to be written into a first or second storage unit according to a data access type and an access management policy according to an embodiment of the invention.

Step S410 of FIG. 4, the decision module 120 determines ₁₅₁ the application access frequency and file size of individual single instruction, access the data amount is consistent with the statistical value of the first type or the second type of memory access rule Take the rules. When the first type of access rule is met, the decision module 120 selects to write the data to the first storage unit 21 as shown in step S420. When the second type of access rule is met, the decision module 120 selects to write the data to the second storage unit 23 as shown in step S430.

For example, in one embodiment, the decision module 120 compares the statistics of the access frequency taken from the application profile database 110 with the high access frequency threshold. If the access frequency reaches the high access frequency threshold, it is determined that the application 15 ₁ may often have high frequency write and read operations, so the decision module 120 selects the first storage unit 21 with high read/write speed. Store data. Instead, the second storage unit 23 stores the data. For example, when a user writes a report or records experimental data using a word processing application or a spreadsheet application, the data is written to the first storage unit because the files are more likely to be read or modified in the near future. 21 can significantly shorten the time the user waits for the file to open.

In another embodiment, the decision module 120 compares the statistical value of the file size taken from the application property database 110 with the large file capacity lower limit. If the file size reaches the upper limit of the large file capacity, the decision module 120 will select to store the data in the second storage unit 23 to prevent the large file from exhausting the available space of the first storage unit 21 too quickly. For example, high-definition movies have file sizes that are billions of bits (GigaByte, GB) and are often less likely to be read or edited by users after storage. If such a file is stored in the first storage unit 21 having a small capacity, the fast reading and writing capability of the first storage unit 21 cannot be quickly utilized to process the data to be subsequently written and read. If the second storage unit 23 having a capacity higher than the first storage unit 21 is stored, the above situation can be avoided, and since the number of times such files are accessed later is less frequent, even the second storage unit 23 The access speed is slower and the difference is not obvious to the user.

In addition, since the flash memory storage unit is a unit that erases physical blocks as data, and uses a physical page as a unit for data writing. Each physical block will include several physical pages. Therefore, when the decision module 120 selects to write data to the first storage unit 21, it can further determine whether the data to be written exceeds the capacity of one physical block, and if so, writes in a whole block (Entire Block) Write, EBW) way to write data in a continuous manner to improve data writing speed. Conversely, if the data to be written does not exceed the capacity of one physical block, the data is written in a normal manner.

In another embodiment, the decision module 120 determines that the application program 15 ₁ is more commonly used in a random access mode or a continuous access mode according to a statistical value of a single instruction access data amount taken from the application property database 110. For example, in the random access mode, the amount of data accessed per instruction does not exceed 16 kilobytes, so if the single instruction access data amount from the application property database 110 is less than the above threshold. For example, the decision module 120 determines that the application program 15 ₁ uses the random access mode. Therefore, the first storage unit 21 with a faster access speed is selected to store the data to be written by the application program 15 ₁ . On the contrary, the decision module 120 determines that the application program 15 ₁ adopts the continuous access mode. Since the advantage of the high access speed of the first storage unit 21 is not fully displayed in the continuous access mode, the decision module 120 will The second storage unit 23 is selected to store the data to be written by the application 15 ₁ .

It is worth mentioning that the decision module 120 updates the data access type of the corresponding application recorded in the application property database 110 according to the content of the write command received each time. For example, adjust the access frequency, file size, and individual instruction to access individual statistics. Moreover, the decision module 120 periodically corrects the first type of access rules and/or the second type of access rules according to the operational performance of the storage device 20.

For example, if the data access type of the document processing application recorded by the application property database 110 conforms to the first type of access rules, the decision module 120 will select to write the data generated by the document processing application. The first storage unit 21 is entered. However, if a user creates a file with a file processing application that exceeds the upper limit of the large file capacity, After the files created by the document processing application are written into the first storage unit 21 according to the existing rules, the storage device 20 returns the fast-depleted information of the first storage unit 21 to the decision module 120. Based on this, the decision module 120 can know that the current access rule does not apply to the user's usage habits, and then adjusts the content of the access rule to ensure that the file created by the document processing application is written. The storage unit 23 is no longer written to the first storage unit 21. In other words, the storage device 20 actively detects the usage performance of the first storage unit 21 and the second storage unit 23 during the data access process after the data is written by the control module 120. The feedback module 120 is configured to cause the decision module 120 to update the first type of access rules and/or the second type of access rules according to the information reported by the storage device 20 to conform to the usage habits of the individual users.

In the above embodiment, the decision module 120 can determine, according to the information provided by the application property database 110, the data to be written by the application 15 ₁ to be written into the first storage unit 21 or the second storage unit 23. Furthermore, the decision module 120 can be assigned based on application data access patterns for the ₁₅₁ space efficient writing and management files, and file reading and writing as much as possible in a continuous manner, reduce the random access The resulting speed drop, resulting in better read and write performance.

On the other hand, the above operation mode of the decision module 120 can also reduce the fragmented space in the first storage unit 21 using the flash memory as a storage medium, and directly reduce garbage collection (Garbage). Collection, GC) frequency and cost, and effectively suppress the Write Amplification Factor (WAF), drop The Erase count of the memory cells in the first storage unit 21 is such that the lifetime of the first storage unit 21 is prolonged and the reliability thereof is improved.

Moreover, the operation of the decision module 120 can effectively reduce the actual number of writes of the first storage unit 21 due to the same external workload. Moreover, for the first storage unit 21, since it is limited by a small capacity, if the storage space cannot be properly allocated, the shortage of capacity will be quickly encountered, and in order to solve such a dilemma, it is necessary to frequently do the internals. Garbage collection, data movement, and storage of files back to other storage devices, so that reading, writing, and erasing operations are more frequent and more power-hungry, so by writing files efficiently It can also improve its power consumption performance.

In summary, the data writing method and system of the present invention can select different types of storage units according to the data access type of the application to store the data to be written by the application. This ensures that the storage device has made significant advances in terms of read and write performance, reliability, and power consumption, allowing users to enjoy the benefits of different types of storage media.

Although the present invention has been disclosed in the above embodiments, it is not intended to limit the invention, and any one of ordinary skill in the art can make some modifications and refinements without departing from the spirit and scope of the invention. The scope of the invention is defined by the scope of the appended claims.

10‧‧‧Host system

13‧‧‧Operating system

15 ₁ , 15 _n ‧‧‧Applications

20‧‧‧Storage device

21‧‧‧First storage unit

23‧‧‧Second storage unit

100‧‧‧data writing system

110‧‧‧Application Properties Database

120‧‧‧Decision module

130‧‧‧Database building module

S210~S240‧‧‧ steps of the data writing method according to an embodiment of the present invention

S410~S430‧‧‧ Steps for selecting data to be written into the first or second storage unit according to the data access type and the access management policy according to an embodiment of the present invention

1 is a schematic diagram of the use of a data writing system according to an embodiment of the invention.

2 is a flow chart of a method of writing data according to an embodiment of the invention.

3 is a bar graph of an application and an access type according to an embodiment of the invention.

4 is a flow chart of selecting data to be written into a first or second storage unit according to a data access type and an access management policy according to an embodiment of the invention.

S210~S240‧‧‧ steps of the data writing method according to an embodiment of the present invention

Claims (12)

A data writing method for writing a data from a host system to a storage device having a first storage unit and a second storage unit, the method comprising: providing an application property database; receiving and executing An application of the host system to write the data to a write command of the storage device; obtaining a data access type corresponding to the application from the application property database; and accessing the data according to the data access type And the access management policy selects to write the data to the first storage unit or the second storage unit. The data writing method of claim 1, wherein the access management policy includes a first type of access rule corresponding to the first storage unit and a second type corresponding to the second storage unit. Access rules. The method for writing data according to claim 2, wherein the data access type comprises an access frequency, a file size, and a single instruction to access a certain amount of data, and the data is stored according to the data. The fetching mode and the access management policy selecting to write the data to the first storage unit or the second storage unit include: when the access frequency, the file size, and the single instruction access data amount are individual When the statistical value conforms to the first type of access rule, the data is written into the first storage unit; and when the access frequency, the file size, and the single instruction access data amount are respectively compliant with the statistical value Write the data when the second type of access rule The second storage unit. The method for writing data according to claim 3, wherein the step of providing the application property database comprises: collecting the access frequency of the application, the file size, and the single instruction in a specific period of time. Taking a plurality of actual values of the data amount; calculating, according to the actual values, the access frequency, the file size, and the statistical value of the single instruction access data amount; and the access frequency, the file size, and the single The statistical value of the instruction access data amount is recorded in the application property database. The method for writing data according to item 2 of the patent application scope further includes: updating, according to the content of the write instruction, the data access type corresponding to the application recorded in the application property database; and periodically The first type of access rules and/or the second type of access rules are modified in accordance with an operational performance of the storage device. The data writing method of claim 2, wherein the first storage unit is a flash memory storage unit, and the first type of access rule includes at least a high access frequency threshold and a random number. The access mode, wherein the second storage unit is a disk storage unit, and the second type of access rule includes at least a large file capacity lower limit value and a continuous access mode. A data writing system includes: an application property database; and a decision module coupled to the application property database, the decision module is configured to execute an application executed in a host system to write a data Enter When a write command of the storage device is obtained, a data access type corresponding to the application is obtained from the application property database, and the data is selected according to the data access mode and an access management policy. And entering a first storage unit or a second storage unit of the storage device. The data writing system of claim 7, wherein the access management policy includes a first type of access rule corresponding to the first storage unit and a second type corresponding to the second storage unit. Access rules. The data writing system described in claim 8 wherein the data access type includes an access frequency, a file size, and a single instruction access data amount, and the decision module When the access frequency, the file size, and the statistic value of the single instruction access data amount meet the first type of access rule, the data is selected to be written into the first storage unit, and the decision module is in the When the access frequency, the file size, and the statistic value of the single instruction access data amount conform to the second type of access rule, the data is selected to be written into the second storage unit. The data writing system of claim 9 further includes: a database building module coupled to the application property database, the database building module collecting the application during a specific period of time The access frequency, the file size, and the plurality of actual values of the single instruction access data amount, and calculating the access frequency, the file size, and the statistical value of the single instruction access data amount according to the actual values, And recording the statistical value of the access frequency, the file size, and the single instruction access data amount In the application feature database. The data writing system of claim 8, wherein the decision module updates the data access type corresponding to the application recorded in the application property database according to the content of the write command. The first type of access rules and/or the second type of access rules are periodically modified based on an operational performance of the storage device. The data writing system of claim 8, wherein the first storage unit is a flash memory storage unit, and the first type of access rule includes at least a high access frequency threshold and a random number. The access mode, wherein the second storage unit is a disk storage unit, and the second type of access rule includes at least a large file capacity lower limit value and a continuous access mode.