JP7334423B2 - Memory control device, memory control program, and memory control method - Google Patents

Description

The present invention relates to a memory control device, a memory control program, and a memory control method.

In a storage system that redundantly includes a plurality of storage controllers (CMs; Controller Modules), each CM stores data such as log information (also called a system log) about the storage system in memory. Then, these CMs redundantly store the data stored in each memory in one or more backup devices (BUDs) provided in each CM, for example, at a predetermined timing.

Japanese Patent Publication No. 2008-539533 JP 2014-32582 A

In the above-described storage system, each CM writes data to each BUD almost evenly, so there are cases where these BUDs exceed their write lifetimes (for example, write limits) at almost the same timing. In this case, there is a problem of losing BUD redundancy.

In one aspect, the goal is to maintain BUD redundancy.

A memory control device that controls writing of data to be backed up in a storage control device using a plurality of semiconductor memory devices having a limited number of writable times,
When a plurality of semiconductor memory devices are in operation , some of the plurality of semiconductor memory devices to be write-restricted are determined based on the identifiers of the semiconductor memory devices , and By imposing write restrictions on some of the semiconductor memory devices subject to the restriction based on the state related to the life of the device and the importance of the data to be backed up , the semiconductor memory devices subject to the restriction are restricted. a control unit for delaying the time until reaching a failure state later than other semiconductor memory devices other than the part of the semiconductor memory devices subject to the restriction among the plurality of semiconductor memory devices.

According to one embodiment, BUD redundancy can be maintained.

1 is a diagram illustrating the hardware configuration of a storage system as an example of an embodiment; FIG. 2 is a diagram illustrating the hardware configuration of CMs in the storage system as an example of an embodiment; FIG. 3 is a diagram exemplifying the functional configuration of a CM in a storage system as an example of an embodiment; FIG. FIG. 4 is a diagram exemplifying BUD state determination information in a storage system as an example of an embodiment; FIG. 4 is a diagram exemplifying write control information in a storage system as an example of an embodiment; FIG. 4 is a diagram exemplifying importance determination information in a storage system as an example of an embodiment; FIG. 4 is a diagram showing an example of how a BUD storage area is used in a storage system as an example of an embodiment; FIG. 4 is a diagram illustrating write control processing to a BUD in a storage system as an example of an embodiment; FIG. 10 is a diagram illustrating importance determination processing of write target data in the storage system as an example of an embodiment; FIG. 4 is a diagram illustrating lifespan prediction processing in a storage system as an example of an embodiment;

BEST MODE FOR CARRYING OUT THE INVENTION Hereinafter, embodiments of the present invention will be described with reference to the drawings. However, the embodiments described below are merely examples, and there is no intention of excluding various modifications and application of techniques not explicitly described below. For example, this embodiment can be modified in various ways without departing from the spirit of the embodiment. In the drawings used in the following embodiments, parts with the same reference numerals represent the same or similar parts unless otherwise specified.

[1] One Embodiment [1-1] Hardware Configuration Example of Storage System According to One Embodiment FIG. 1 is a diagram illustrating the hardware configuration of a storage system 1 as an example of an embodiment.

As illustrated in FIG. 1, the storage system 1 includes two or more CMs (CM10-0, . , server device 30-m; m is an integer equal to or greater than 0). The storage system 1 also includes a plurality of HDDs (Hard Disk Drives) 50 and an inter-CM communication switch 60 .

The plurality of CM10-0, . . . , CM10-n are also represented as CM#0, . Hereafter, CM 10-0 to 10-n will be referred to as CM 10 unless otherwise distinguished. The numbers (0, . . . , n) following “#” are also called identifiers (IDs) and are used to identify the CM 10 .

These CMs 10 are storage control devices and control input/output (read and write) of data to each HDD 50 . The CM 10 provides the storage area of the HDD 50 to the server device 30 .

The one or more server devices 30-0, . . . , 30-m are also represented as server devices #0, . Hereinafter, the server devices 30-0 to 30-m are referred to as the server device 30 unless otherwise distinguished. Note that the numbers (0, . . . , m) following “#” are also called identifiers and are used to specify the server device 30 . Also, FIG. 1 illustrates a case where one CM 10 (eg, CM#0) is connected to one server device 30 (eg, server device #0), but the present invention is not limited to this.

This server device 30 is a computer having a server function, and transmits data access requests for the BUD 20 and HDD 50 to each CM 10 . The server device 30 transmits, to the CM 10 to which it is connected, a data access request to the BUD 20 and the HDD 50 provided in the CM 10 other than the CM 10 (subordinate to the CM 10 other than the CM 10 to which the server device 30 is connected). may

The inter-CM communication switch 60 receives a communication request from the CM 10, detects, for example, the identifier of the communication destination CM 10 included in the request, and establishes a path connection between the request source CM 10 and the communication destination CM 10. do The inter-CM communication switch 60 also transfers data read requests and write requests (data access requests) from one CM 10 to the other CM 10 .

Each CM 10 comprises one or more BUDs 20, which are storage devices. FIG. 1 illustrates a case where each CM 10 has two BUDs 20 . As shown in FIG. 1, the storage system 1 includes BUD#0 (20-0) to BUD#m (20-m; m is an integer equal to or greater than 0). BUD20-0, . . . , BUD20-m are also expressed as BUD#0, . Hereinafter, BUD20-0 to 20-m will be referred to as BUD20 unless otherwise distinguished. The numbers (0, . . . , m) following “#” are also called identifiers and are used to identify the BUD 20 .

The BUD 20 is, for example, an SSD (Solid State Drive) and has a non-volatile semiconductor memory as a memory element. It should be noted that the BUD 20 is not limited to an SSD and can be modified in various ways.

The HDD 50 is a magnetic disk device. FIG. 1 shows an example in which each CM 10 has two or more HDDs 50 .

In this storage system 1, a RAID (Redundant Array of Inexpensive Disks) may be configured by combining a plurality of these HDDs 50, which are physical drives, to configure a virtual logical drive.

Although FIG. 1 shows an example in which each CM 10 includes two BUDs 20, the present invention is not limited to this. CM 10 may comprise one or more BUDs 20 . Also, the number of HDDs 50 connected to each CM 10 can be changed as appropriate.

Also, the CM 10 may be connected to, for example, a management terminal (not shown) used by a system administrator via a network (not shown).

[1-2] Hardware Configuration Example of CM in Storage System According to One Embodiment FIG. 2 is a diagram illustrating the hardware configuration of the CM 10 (CM#0) in the storage system 1 as an example of the embodiment.

The hardware configuration of the CM 10 shown in FIG. 1 will be described using FIG. In this embodiment, each of the CMs 10 illustrated in FIG. 1 has the hardware configuration illustrated in FIG.

The CM 10 may include, for example, a CPU (Central Processing Unit) 11 , a storage section 12 , a memory 13 , an IF (Interface) section 14 , an input/output section 15 and a display section 16 .

The CPU 11 executes an OS (Operating System) and a control program 90 (described later) stored in the storage unit 12 described later, and for example, according to an instruction from the server device 30, for example, an arithmetic processing device (processor) that controls the BUD 20 ).

The storage unit 12 is an example of hardware that stores various data, programs, and the like. For example, the storage unit 12 may be used as a secondary storage device for the CM 10, and may store an OS, firmware, programs such as applications, and various data. The storage unit 12 may be, for example, an HDD, an SSD, an SCM (Storage Class Memory), or another storage device. The storage unit 12 may store a program (a control program 90 to be described later) that implements all or part of various functions of the CM 10 .

The memory 13 is an example of hardware that stores various data, programs, and the like. Examples of the memory 13 include volatile memory such as RAM (Random Access Memory) and nonvolatile memory such as flash memory, SCM, ROM (Read Only Memory). The memory 13 may also store the control program 90 . The RAM of memory 13 may be used as primary storage memory or working memory.

The IF unit 14 is an example of a communication interface that controls connections and communications with other CMs 10, the server device 30, and a management device (not shown). For example, the IF section 14 may have an adapter (port) for connecting a management device (not shown). The IF unit 14 may include, for example, a PCIe (Peripheral Component Interconnect Express) interface or an optical communication interface. Note that the control program 90 may be downloaded via the IF unit 14 from a network (not shown).

The input/output unit 15 may include, for example, at least one of a mouse, a keyboard, a touch panel, and an input device such as an operation button.

The display unit 16 may include, for example, at least one of a display, a projector, a speaker, and an output device such as a printer.

The hardware configuration of the CM 10 described above is an example. Therefore, the hardware within the CM 10 may be increased or decreased (for example, addition or omission of arbitrary blocks), division, integration in arbitrary combinations, or the like may be performed as appropriate.

[1-3] Example of Functional Configuration in Storage System in One Embodiment The functional configuration of the CM 10 according to this storage system will be described with reference to FIG.

FIG. 3 is a diagram illustrating the functional configuration of the CM 10 as an example of the embodiment shown in FIG. 1. As shown in FIG. In this embodiment, each of the CMs 10 illustrated in FIG. 1 has the functional configuration illustrated in FIG.

As illustrated in FIG. 3, the CM 10 illustratively includes an inter-CM communication unit 41, a write restriction BUD determination unit 42, an importance determination unit 43, a status acquisition unit 44, a write mode determination unit 45, a state determination unit 46, Also, a write execution unit 47 may be provided.

The inter-CM communication unit 41 communicates with (the inter-CM communication unit 41 of) another CM 10 and obtains the operating status of each CM 10 that constitutes the storage system 1 and the operating status of the BUD 20 provided in each CM. In this embodiment, the operating status of the CM 10 is, for example, information indicating whether the CM 10 is operating or out of order, and the operating status of the BUD 20 is, for example, whether the BUD 20 is operating or out of order. It is information indicating whether

The write limit BUD determination unit 42 determines a write limit BUD 20 that limits (reduces) the amount of data to be written among the plurality of BUDs 2020 provided in the storage system 1 .

This write restriction BUD determination unit 42 determines the number of BUDs 20 in operation in the storage system 1 .

When the write restriction BUD determination unit 42 determines that two or more BUDs 20 are in operation in the storage system 1, only one of the plurality of BUDs 20 is designated as a normal write BUD for which data writing is not restricted. Determine BUD20. For example, the write-restricted BUD determination unit 42 determines the BUD 20 (BUD#0) having the smallest identifier (0 in this example) among the BUDs 20 in operation as the normal write BUD 20 .

Note that even when three or more BUDs 20 are in operation (determined), it is desirable to determine only one of these BUDs 20 as the normal write BUD 20 . This is to reduce the amount of writing to the BUD 20 in the entire storage system 1 and extend the life of the BUD 20 by limiting the number of BUDs to which data writing is not restricted to one.

Also, the write restriction BUD determination unit 42 determines the write restriction BUD 20 which is a BUD that restricts (reduces) the contents of writing to the BUD 20 . In this embodiment, the write restriction BUD determination unit 42 determines the BUDs 20 other than the normal write BUD 20 (BUDs #1 to #m in this example) as the write restriction BUDs 20 .

By limiting data (amount of data) to be written in the write limit BUD 20, the write limit BUD 20 delays the time required for the cumulative write amount to reach the limit value and extends the life of the write limit BUD 20 compared to the normal write limit BUD 20. - 特許庁As a result, the write-restricted BUD 20 can take longer to replace or fail than the normal write BUD 20 . In this way, by providing the normal write BUD 20 and the write restriction BUD 20 in the plurality of BUDs 20 that realize the redundant configuration provided in the storage system 1, it is possible to create a difference in the timing of failure occurrence of the BUD 20 and maintenance work. can be done. That is, the redundancy of the BUD 20 can be maintained.

The importance determination unit 43 determines the importance of data to be written to the BUD 20, that is, data to be written. The importance determination unit 43 selects, for example, one of three types of importance of "H" (high), "M" (middle), and "L" (low) for the data to be written. to determine (determine). A specific process for determining the degree of importance will be described later. Note that the data to be written to the BUD 20 in this embodiment is, for example, data on the memory 13 , such as log information and backup data of the storage system 1 .

Further, as will be described later, the importance determination unit 43 determines the importance of the write target data based on the write destination address to the BUD 20 and the type of the write target data. Therefore, the importance determination unit 43 acquires the write destination address to the BUD 20 and the type of the write target data.

The types of this data are, for example, system log, failure determination/abnormality detection log, write mode transition log, environmental temperature/voltage, and external connection change. The importance determination unit 43 may, for example, acquire the type of the data based on a flag included in the write target data. Further, the importance determination unit 43 may determine the type of data based on specific data included in the write data, and can be implemented in various modifications. Also, since the method of acquiring the type of data to be written is well known, the description thereof is omitted here.

The status acquisition unit 44 stores and manages the operation time of each BUD 20 provided in the CM 10 to which it belongs, for example, in the storage unit 12 . The status acquisition unit 44 can acquire the operating time of each BUD 20 stored in the storage unit 12 as appropriate.

The status acquisition unit 44 also accesses each BUD 20 provided in the CM 10 to which it belongs, and acquires the cumulative write amount (for example, disk usage) of the BUD 20 . Then, the status acquisition unit 44 stores and manages the acquired cumulative write amount of the BUD 20 in the storage unit 12, for example. Incidentally, in the present embodiment, the status is the cumulative write amount of each BUD 20 and the operating time of each BUD 20 .

The write mode determination unit 45 determines a write mode indicating how to limit the amount of data to be written in the write limit BUD 20 . The write mode determination unit 45 determines the write mode from among three types, for example, "normal mode", "prediction avoidance mode", and "failure avoidance mode". These write modes are indicators for restricting writes to the write restriction BUD 20 . Therefore, when the write mode for the write-restricted BUD 20 is changed by the write mode determination unit 45, the content (type) of data written to the write-restricted BUD 20 is changed. Therefore, the amount of data written to the BUD 20 is changed. It is assumed that the write mode is not set for the normal write BUD 20 .

Each of these write modes will be described. “Normal mode” is a mode in which all data (types) are written without restricting data writing to the BUD 20 . If there is still a margin before the cumulative write amount of the BUD 20 reaches the limit value of the write amount of the BUD 20 at the design stage, the write mode determination unit 45 sets the "normal mode" and sets a special limit. Then, the BUD 20 is written to. A specific method of writing by setting this "normal mode" will be described later. Note that the same data as in this "normal mode" is written to the normal write BUD 20. FIG.

On the other hand, the “prediction avoidance mode” and the “failure avoidance mode” are examples of modes (write restriction modes) that restrict writing of data to the BUD 20 .

The “failure avoidance mode” is a mode in which, of the data to be written, the types of data less than those in the “prediction avoidance mode” are written to the BUD 20 . The "prediction avoidance mode" is a mode in which less types of data than in the "normal mode" are written, but more types of data than in the "failure avoidance mode" are written. A specific method of setting the "failure avoidance mode" or "prediction avoidance mode" for writing will be described later. The amount of data to be written decreases in the order of "normal mode", "prediction avoidance mode", and "failure avoidance mode".

A state determination unit 46 determines the state of the BUD 20 . The state determination 46 classifies the state of the BUD 20 into four types of "normal state", "predicted failure state", "predicted failure sign state", and "failure state" based on the operating time and accumulated write amount of the BUD 20. Judgment (decision) is made from among the states of

The state determination unit 46 determines the state of the BUD 20 by referring to the BUD state determination information 70 illustrated in FIG.

FIG. 4 is a diagram illustrating an example of an embodiment and the BUD state determination information 70 in the storage system 1. As illustrated in FIG.

The BUD state determination information 70 illustrated in FIG. 4 indicates the relationship between two parameters, namely, the operation time of the BUD 20 (horizontal axis) and the accumulated write amount of the BUD 20 (vertical axis), and the state of the BUD 20 .

The "design life" on the horizontal axis of FIG. 4 indicates the write life of the BUD 20 at the design stage. When the operation time of the BUD 20 reaches the write life, it is known that the time to replace the BUD 20 has come, for example.

Also, the "limit value" on the vertical axis of FIG. 4 indicates, for example, the limit value of the write amount of the BUD 20 at the design stage. It should be noted that writing data exceeding this limit value in the BUD 20 is not guaranteed.

Further, as illustrated in FIG. 4, the state (operating state) of the BUD 20 is set to "normal", " It is classified into four states of "failure prediction state", "failure sign prediction state", and "failure state".

Specifically, the BUD state determination information 70 illustrated in FIG. 4 includes a linear function straight line L1 rising to the right and a linear function straight line L2 rising to the right located above the linear function straight line L1. A region below the linear function line L1 and sandwiched between the linear function line L1 and the horizontal axis corresponds to the "normal state."

Further, the area above the linear function line L2 and sandwiched between the linear function line L2 and the vertical axis corresponds to the "failure prediction state". Further, the area sandwiched between the linear function straight lines L1 and L2 at the upper position of the linear function straight line L1 and the lower position of the linear function straight line L2 corresponds to the "predicted failure sign state". A method for determining each of these states will be described below.

The state determination unit 46 determines the state of the BUD 20 by referring to the BUD state determination information 70 illustrated in FIG. The state determination unit 46 determines the state of the BUD 20 according to which region in the BUD state determination information 70 the combination of the accumulated write amount and operating time of the BUD 20 belongs to.

A method of determining the state of the BUD 20 by the state determination unit 46 will be described using an example in which the operation time of the BUD 20 is x years (x is 0 or more, but x (years) is less than the design life).

In the example shown in FIG. 4, when the cumulative write amount is less than a (GB), for example, the state determination unit 46 determines that the BUD 20 is in a "normal state." However, a is 0 or more and less than b described later.

Further, when the cumulative write amount is, for example, c (GB) or more, the state determination unit 46 determines that the state of the BUD 20 is "failure state", that is, the write life has been reached. However, c is 0 or more and less than the limit value.

Further, as illustrated in FIG. 4, when the cumulative write amount is, for example, b (GB) or more and less than c (GB), the state determination unit 46 determines that the state of the BUD 20 is the "failure prediction state", that is, , is determined to be the state before reaching the above-described "failure state". However, b is 0 or more and less than c. Thus, when the state determination unit 46 determines that the state of the BUD 20 is the "failure prediction state", it can be predicted that the life of the BUD 20 will soon reach the write life.

Further, when the cumulative write amount is, for example, a (GB) or more and less than b (GB), the state determination unit 46 determines that the state of the BUD 20 is not the “failure sign prediction state”, that is, the “normal state”. And, it is determined that the state is before reaching the "failure prediction state" described above. In this way, when the state determination unit 46 determines that the state of the BUD 20 is the "failure predictive state", the life of the BUD 20 does not immediately reach the write life, but the state is determined to be normal. I can predict that it will pass.

In this manner, the state determination unit 46 refers to the BUD state determination information 70 illustrated in FIG. 4 to determine the state of each BUD 20 provided in the CM 10 to which it belongs. It should be noted that the state of the BUD 20 illustrated in FIG. 4 is merely one aspect, and can be modified in various ways without departing from the scope of the invention.

Also, the BUD state determination information 70 illustrated in FIG. 4 is merely one aspect, and can be modified in various ways without departing from the gist thereof. That is, the BUD state determination information 70 illustrated in FIG. 4 merely represents the relationship between the operating time of the BUD 20, the accumulated amount of writing, and the state of the BUD 20 as a graph for convenience, and can be modified as appropriate. .

The write execution unit 47 writes data to the BUD 20 . The write execution unit 47 refers to the write control information 72 (see FIG. 5) and determines a write control pattern based on the write mode and the importance of the data to be written. The write control pattern indicates whether the writing of data to be written to the write-restricted BUD 20 is performed by normal writing or by inhibiting writing. This write control information 72 will be described later with reference to FIG. The write execution unit 47 writes the write target data to the write restriction BUD 20 according to the determined write control pattern.

FIG. 5 is a diagram exemplifying the write control information 72. As shown in FIG. The write control information 72 illustrated in FIG. 5 indicates that whether or not the write target data is written to the write restriction BUD 20 differs depending on the write mode. Furthermore, the write control information 72 illustrated in FIG. 5 indicates that whether writing to the write restriction BUD 20 is performed or not depends on the importance of the data to be written.

When the write mode is the "normal mode", the write execution unit 47 writes the write target data to the write restriction BUD 20 regardless of the importance of the write target data. When the write mode is the "normal mode", it can be judged that there is a margin before the cumulative write amount of the write limit BUD 20 reaches the limit value, so writing to the write limit BUD 20 is performed without setting a limit.

When the write mode is the "failure avoidance mode", the write execution unit 47 writes only the data whose importance is "H" to the write restriction BUD 20 (performs normal writing). Here, writing to the write limit BUD 20 will be described with reference to FIG.

FIG. 6 is a diagram exemplifying the importance determination information 71 in the storage system 1 according to one embodiment, and FIG.

The storage areas of the BUD 20 illustrated in FIG. 7 exemplarily include a firmware storage area 201, a log information storage area 202, a backup area 203, and a free area 204. FIG. In addition, in this embodiment, each of the storage areas of the plurality of BUDs 20 illustrated in FIG. 1 is provided with each area illustrated in FIG. 7 .

The firmware storage area 201 is an area in which firmware executed by the CPU 11 to implement the functions of the CM 10 is stored.

The log information storage area 202 is an area in which log information of data stored in the memory 13 of the CM 10 is stored (backed up).

The backup area 203 is an area in which data to be backed up among data stored in the memory 13 of the CM 10, excluding log information, is stored. Data to be backed up excluding this log information is called backup data.

A free area 204 is a buffer area in which no data is stored (unused).

The above-described firmware storage area 201, log information storage area 202, backup area 203, and free area 204 are divided based on the address (start address) of the storage area of the BUD 20. FIG. Therefore, by referring to the write destination address included in the write request to the BUD 20, the write target data of the write request can be specified.

The importance determination information 71 illustrated in FIG. 6 shows the write destination area, the type of data to be written, and the importance of the data to be written in association with each other.

Specifically, the importance determination information 71 illustrated in FIG. 6 indicates that the importance of data whose write destination area is the firmware storage area 201 or the backup area 203 is "H" (see arrow R1). Also, it is indicated that the data whose write destination area is the log information storage area 202 and whose type of data to be written is any one of the failure determination log, the abnormality detection log, and the write mode transition log has the importance level "M". (see arrow R2). Furthermore, it is determined that the data whose write destination area is the log information storage area 202 and whose type of data to be written is any of environmental temperature, voltage, and logs indicating changes in external connection have the importance level "L". (see arrow R3).

Therefore, for example, in the write control information 72 illustrated in FIG. 5 , in the case of the “failure avoidance mode”, the write execution unit 47 writes data of importance “H” to the write restriction BUD 20 . That is, the write execution unit 47 refers to the importance determination information 71 and writes only data whose write destination address indicates the firmware storage area 201 (see FIG. 7) or the backup area 203 (see FIG. 7). Write to limit BUD 20.

In this "failure avoidance mode", as shown in FIG. 5, the write execution unit 47 does not write to the write-restricted BUD 20 for data whose importance levels are "M" and "L". Suppress (do not write). That is, in the case of the "failure avoidance mode", the write execution unit 47 writes the data (data with the degree of importance "M") whose write target data types are the failure determination log, the abnormality detection log, and the write mode transition log. inhibits writing to the write-restricted BUD 20 .

Similarly, in this "failure avoidance mode", the write execution unit 47 writes data whose type of data to be written is a log indicating changes in environmental temperature, voltage, and external connection (data with importance level "L"). ) is also inhibited from writing to the write-restricted BUD 20 . This is because when the write mode is the "failure avoidance mode", the "failure state" (write life) may be reached soon, so only data with high importance is quickly written to the write limit BUD 20. This is to allow

On the other hand, in the write control information 72 illustrated in FIG. 5 , when the write mode is the “prediction avoidance mode”, the write execution unit 47 writes only the data with the importance levels “H” and “M” of the data to be written. Write to the limit BUD 20 (normal write). Here, referring to FIG. 6, in the case of the "prediction avoidance mode", the write execution unit 47 writes data indicating that the write destination address is the firmware storage area 201 or the backup area 203 (importance level "H"). data) to the write limit BUD 20. The write execution unit 47 also writes to the write restriction BUD 20 data (data with importance level "M") whose types of data to be written are the failure determination log, the abnormality detection log, and the write mode transition log.

Then, in the "prediction avoidance mode", as shown in FIG. 5, the write executing unit 47 suppresses writing to the write-restricted BUD 20 for the data with the importance level of "L" as the data to be written (write is not performed). In other words, the write execution unit 47 writes to the write restriction BUD 20 the type of data to be written that is a log indicating changes in environmental temperature, voltage, and external connection (data with a degree of importance “L”). Deter. This is because when the write mode is "prediction avoidance mode", it can be judged that there is more time to reach the failure state than in "failure avoidance mode", so a looser write limit than in "failure avoidance mode" is set. is.

[1-4] Write control processing to the BUD in the storage system in one embodiment As an example of the embodiment configured as described above, the write control processing to the BUD 20 in the storage system 1 is described with reference to FIGS. 7, FIG. 9, and FIG. 10, description will be made according to the flowchart (steps S1 to S8) shown in FIG.

FIG. 8 is a flowchart for explaining write control processing to the BUD 20 in the storage system 1 according to one embodiment.

FIG. 9 is a flowchart (steps T1 to T6) for explaining the process of determining the importance of write target data in the storage system 1 according to one embodiment.

FIG. 10 is a flowchart (steps Q1 to Q5) for explaining the process of predicting the life in the storage system 1 according to one embodiment.

Note that the writing process to the BUD 20 shown in FIG. 8 may be autonomously started by the CM 10, for example, each time a predetermined interval elapses, but is not limited to this. For example, in the processing shown in FIG. 8, when the operation of the storage system 1 is stopped due to loss of external power, the CM 10 sends an I/O (Input/Output) instruction to the server device 30 to store the data on the memory 13 in the BUD 20. may be triggered by receiving from

In this embodiment, the CM 10 autonomously stores the log information and backup data of the storage system 1 in its own memory 13 at a predetermined timing, but the present invention is not limited to this. For example, when the CM 10 receives an I/O instruction to store the log information and backup data of the storage system 1 in the memory 13 from the server device 30, the information may be stored in its own memory 13. good.

In step S1, the inter-CM communication unit 41 of the CM 10 communicates with the other CM 10 (the inter-CM communication unit 41 of the other CM 10) to determine the operating status of each CM 10 that constitutes the storage system 1 and the information provided by each CM. The operating status of the BUD 20 is acquired.

The inter-CM communication unit 41 acquires, as the operating status of the CM 10, for example, whether each of the plurality of CMs 10 constituting the storage system 1 is in operation or out of order. In addition, the inter-CM communication unit 41 acquires, as the operation status of the BUD 20, whether each BUD 20 provided in each CM 10 is in operation or out of order, for example.

In subsequent step S2, the write restriction BUD determination unit 42 of the CM 10 determines how many BUDs 20 are in operation in the storage system 1 based on the operating status of the BUDs 20 obtained in step S1.

When the write-restricted BUD determination unit 42 determines that two or more BUDs 20 are in operation (“two or more” route in step S2), the process proceeds to step S3.

In subsequent step S3, the write-restricted BUD determination unit 42 determines the BUD 20 with the smallest identifier among the operating BUDs 20 obtained in step S1 to be the normal write BUD 20. FIG.

Also, in this step S3, the write-restricted BUD determination unit 42 determines the BUDs 20 other than the normal write BUDs 20 among the operating BUDs 20 obtained in the above-described step S1 to be the write-restricted BUDs 20. FIG.

For example, in the storage system 1 illustrated in FIG. 1, the CM#0 to CM#n are all in operation, and the BUD#0 to BUD#m included in these CMs 10 are all in operation. do. In this case, the write-restricted BUD determination unit 42 determines BUD#0, which is the BUD 20 with the smallest identifier, as the normal write BUD 20, and determines BUD#1 to BUD#m (other than BUD#0) as the write-restricted BUD 20. .

In subsequent step S4, the importance determination unit 43 determines the importance of the data to be written.

The details of the processing in step S4 will be specifically described according to the flowchart (steps T1 to T6) shown in FIG.

At step T1 shown in FIG. 9, the importance level determination unit 43 acquires a write destination address for the BUD 20. FIG.

In this step T1, when the CM 10 autonomously stores the log information and backup data of the storage system 1 in its own BUD 20, the CM 10 itself may acquire the write destination address of the BUD 20 specified by itself. Further, when the server device 30 instructs the CM 10 to store the above information in the BUD 20, the CM 10 may acquire the write destination address included in the instruction.

In subsequent step T2, the importance determination unit 43 acquires the type of data to be written to the BUD 20. FIG. In this step T2, the importance determination unit 43 may acquire the type of the data based on the flag included in the data to be written, as described above.

In the subsequent step T3, the importance determination unit 43 converts the importance determination information 71 illustrated in FIG. 6 based on the write destination address acquired in step T1 and the type of write target data acquired in step T2. is used to determine the importance of the data to be written.

The importance determination unit 43 determines whether the write destination address acquired in step T1 is included in any of the firmware storage area 201, the log information storage area 202, and the backup area 203 among the areas of the BUD 20 illustrated in FIG. to judge whether

Further, the importance determination unit 43 determines whether the type of the write target data acquired in step T2 is any of the failure determination log, the abnormality detection log, and the write mode transition log illustrated in FIG. to decide. These failure determination log, abnormality detection log, and write mode transition log are included in the log information.

Further, the importance determination unit 43 determines that the type of the data to be written acquired in step T2 is environmental temperature (environmental temperature), voltage, and change in external connection (for example, cable disconnection), which are illustrated in FIG. Determine if it is one of the logs shown. Logs indicating the temperature and voltage of these environments and logs indicating changes in external connections are included in the log information of the storage system 1 .

As a result of referring to the importance determination information 71, if the write destination address acquired in step T1 indicates the firmware storage area 201 or the backup area 203, the importance determination unit 43 sets the importance of the data to be written to "H" ( high) (see step T4). In this way, when the write destination address indicates the firmware storage area 201 or the backup area 203, the importance determination unit 43 determines that the importance is "H" (high) regardless of the type of data to be written. (see arrow "R1" in FIG. 6).

As a result of referring to the importance determination information 71, if the write destination address acquired in step T1 indicates the log information storage area 202, the importance determination unit 43 determines the type of the write target data acquired in step T2. is used to determine importance. Specifically, when the type of data to be written is any one of a failure determination log, an abnormality detection log, and a write mode transition log, the importance determination unit 43 sets the importance of the data to be written to "M (medium) (step T5, see arrow "R2" in FIG. 6).

On the other hand, when the type of data to be written is any one of environmental temperature, voltage, and log indicating changes in external connection, the importance determination unit 43 sets the importance of the data to be written to "L". (low) (step T6, see arrow "R3" in FIG. 6).

In this way, when the write destination address indicates the log information storage area 202, the importance determination unit 43 sets the importance of the write target data to "M" or "L" according to the type of the write target data. I judge. After that, the process ends.

Next, returning to the description of FIG. 8, in step S4, through steps T1 to T6 in FIG. , and "L".

In subsequent step S5, the write mode determination unit 45 of the CM 10 acquires the status from the status acquisition unit 44 and determines the write mode based on the acquired status. The status here is the accumulated write amount of each BUD 20 and the operating time of each BUD 20 .

The details of the processing in step S5 will be specifically described according to the flowchart (steps Q1 to Q5) shown in FIG.

In step Q1 shown in FIG. 10, the write mode determination unit 45 acquires the cumulative write amount of each BUD 20 and the operating time of each BUD 20 from the status acquisition unit 44 as the status.

Specifically, the status acquisition unit 44 acquires the cumulative write amount (for example, disk usage) of each BUD 20 provided in the CM 10 to which the status acquisition unit 44 belongs.

In addition, the status acquisition unit 44 acquires the operation time of each BUD 20 provided in the CM 10 to which the status acquisition unit 44 belongs (for example, the elapsed time from the time of new introduction). In this embodiment, the status acquisition unit 44 stores and manages the operation time of each BUD 20 provided in the CM 10 to which the status acquisition unit 44 belongs in the storage unit 12. In this step Q1, the status acquisition unit 44 manages The operating time of each BUD 20 is referenced.

Note that the operating time of each BUD 20 may be managed by another functional configuration unit (for example, the state determination unit 46), or may be managed by the server device 30. FIG. The status acquisition unit 44 may acquire the operating time of each BUD 20 from these other functional configuration units, the server device 30, and the like.

In subsequent step Q2, the state determination unit 46 refers to the BUD state determination information 70 illustrated in FIG. 4 to determine the state of each BUD 20 provided in the CM 10 to which it belongs.

In this manner, in the process shown in step Q2 of FIG. 10, the state determination unit 46 refers to the BUD state determination information 70 to determine the state of each BUD 20 provided in the CM 10 to which it belongs.

Next, returning to the description of FIG. 10, in the process shown in step Q2, when the state determination unit 46 determines that the state of the BUD 20 is "faulty" ("normal state" route in step Q2), the process proceeds to step Q3. transition to Then, in step Q3, the write mode determination unit 45 determines the write mode to be "normal mode".

Further, when the state determination unit 46 determines that the state of the BUD 20 is the "failure portent prediction state" ("failure portent prediction state" route in step Q2), the process proceeds to step Q4. Then, in step Q4, the write mode determination unit 45 determines the write mode to be the "prediction avoidance mode".

On the other hand, when the state determination unit 46 determines that the state of the BUD 20 is the "predicted failure state" ("predicted failure state" route in step Q2), the process proceeds to step Q5. Then, in step Q5, the write mode determination unit 45 determines the write mode to be the "failure avoidance mode". Then, the process ends.

As described above, in the "failure avoidance mode", more types of write data are restricted than in the "prediction avoidance mode". This is because the write life is reached earlier than in the "prediction avoidance mode".

Next, returning to the description of FIG. 8, in step S5, the write mode is determined according to FIG. 10 (steps Q1 to Q5) described above. This selects the write mode for the write limit BUD20 determined in step S3. Therefore, in this embodiment, the normal write BUD 20 is normally written without any particular limitation.

In subsequent step S6, the write execution unit 47 instructs each BUD 20 to write. Here, the write-restricted BUD 20 determined in step S3 described above is written into the BUD 20 under the write mode determined in step S5 described above. As described above, in the present embodiment, the write executing unit 47 performs normal writing without any particular restrictions on the normal write BUD 20 determined in step S3.

Here, based on the write mode determined in step S5 described above and the importance of the data to be written determined in step S4 described above, the write execution unit 47 assigns the data to be written to the write restriction BUD 20. Decide whether to write normally or suppress. The write control information 72 exemplified in FIG. 5, for example, is referred to when determining whether to perform the writing as usual or to suppress the writing (do not perform the writing).

In step S6, the write execution unit 47 refers to the write control information 72, and if the write mode determined in step S5 is the "normal mode", regardless of the importance of the data to be written, Write the data to the write restricted BUD 20. Here, when the write mode is the "normal mode", it can be determined that there is a margin before the cumulative write amount of the write limit BUD 20 reaches the limit value, so writing to the write limit BUD 20 is performed without setting a limit.

Further, when the write mode determined in step S5 is the "failure avoidance mode", the write execution unit 47 writes only the data with the importance of "H" to the write limit BUD 20 (normal write ). Here, referring to FIG. 6, in the case of the “failure avoidance mode”, only data whose write destination address indicates the firmware storage area 201 or the backup area 203 is written to the write restriction BUD 20 .

Then, the write execution unit 47 suppresses (does not write) the data with the importance levels of "M" and "L" to be written to the write-restricted BUD 20. FIG. Here, referring to FIG. 6, in the case of the "failure avoidance mode", the types of data to be written are the failure determination log, the abnormality detection log, and the write mode transition log (data with the degree of importance "M"). As for, writing to the write restriction BUD 20 is suppressed.

Similarly, writing to the write restriction BUD 20 is also inhibited for data whose type of data to be written is a log indicating changes in environmental temperature, voltage, and external connection (data with importance level "L"). This is because if the write mode is the "failure avoidance mode", the "failure state" (write life) may soon be reached, so only data with high importance is written to the write limit BUD 20. is.

Further, when the write mode determined in step S5 is the "prediction avoidance mode", the write execution unit 47 sets only the data whose importance levels are "H" and "M" to the write restriction BUD 20. Write (perform normal writing). Here, referring to FIG. 6, in the case of the "prediction avoidance mode", the write destination address is the data indicating the firmware storage area 201 or the backup area 203 (data with importance level "H"). write to At the same time, data whose types of data to be written are the failure determination log, the abnormality detection log, and the write mode transition log (data with importance level “M”) are also written to the write restriction BUD 20 .

Then, the write execution unit 47 suppresses (does not write) the data of which the importance level of the data to be written is "L". In other words, the write execution unit 47 writes to the write restriction BUD 20 the type of data to be written that is a log indicating changes in environmental temperature, voltage, and external connection (data with a degree of importance “L”). Deter.

In this step S6, if there are a plurality of write-restricted BUDs 20, the write execution unit 47 similarly writes to each of these write-restricted BUDs 20. FIG.

When the write restriction BUD determination unit 42 of the CM 10 determines that one BUD 20 is in operation in step S2 described above (“1 unit” route in step S2), the process proceeds to step S7.

In the following step S7, the state determination unit 46 determines to write normally to one BUD 20 in operation, and the process proceeds to step S6.

After it is determined that two or more BUDs 20 are in operation after steps S1 to S6, it is assumed that as a result of returning to step S2, the number of BUDs 20 in operation is reduced to one. In this case, in step S7, the previously set write restriction is canceled, and writing is normally performed for one BUD 20 in operation.

In the subsequent step S6, the write execution unit 47 performs normal writing to one BUD 20 that is in operation. That is, the write execution unit 47 performs writing as usual without any particular restrictions.

On the other hand, in step S2, when the write-restricted BUD determination unit 42 of the CM 10 determines that there is no BUD 20 in operation (there are 0 BUDs 20) (“0” route in step S2), the process proceeds to step S8. .

In the following step S8, the write execution unit 47 does not perform writing because there is no BUD 20 to be written (write addition). Then the process ends.

As described above, when a plurality of BUDs 20 are in operation after steps S1 to S3 described above, a write restriction BUD 20 and a normal write BUD 20 are provided (step S3). The normal write BUD 20 is written with the same amount of data as the normal write, while the write limit BUD 20 is made to write less data than the normal write BUD 20 . This prevents the write-restricted BUD 20 from reaching a failure state almost at the same time that the normal write BUD 20 reaches a failure state. As a result, the times until the plurality of BUDs 20 reach the failure state can be staggered (uneven).

Further, through the above steps S4 to S6, the write processing mode is determined (step S5) based on the state of the write restriction BUD 20 (step S5) and the importance of the data to be written (step S5), and the data is restricted. Then, write is executed (step S6). Thereby, writing can be controlled according to the state of the BUD 20, that is, the margin until the BUD 20 reaches the failure state.

[2] Effect As described above, in the storage system 1 according to one embodiment, when a plurality of BUDs 20 are operating, the write restriction BUD 20 and the normal write BUD 20 are provided respectively. As a result, it is possible to stagger (uneven) the time to reach the failure state among the plurality of BUDs 20 .

Therefore, it is possible to avoid replacement or failure of a plurality of BUDs 20 at substantially the same timing, and maintain redundancy of the BUDs 20 .

Further, in the storage system 1 according to one embodiment, the write processing mode is set based on the state of the write restriction BUD 20 and the importance of the data to be written to restrict writing. As a result, the closer the BUD 20 is to the failure state, the more the amount of data to be written is restricted, while the more important data is continuously written, so that the more important data can be reliably backed up. can improve reliability.

Also, in the storage system 1 according to one embodiment, the data stored in the memory 13, which is a volatile memory, can be backed up in the BUD 20. FIG.

In addition, in the storage system 1 according to one embodiment, the operating state of the BUD 20 is determined based not only on the accumulated write amount of the BUD 20 but also on the operating time of the BUD 20 . As a result, rather than simply determining the operation status of the BUD 20 based on the cumulative amount of data written in the BUD 20, it is possible to consider a cause of failure of the BUD 20, such as deterioration over time, thereby improving reliability. can.

In addition, in the storage system 1 according to one embodiment, not only the write destination address but also the type of the write target data is considered to determine the importance of the write target data. As a result, even when the write destination address is the log information storage area 202, different degrees of importance can be set according to the types of log information stored in the area. It is possible to judge that the importance is moderate.

[3] Others The technology according to the above-described embodiment and modifications can be modified and changed as follows.

In the above-described embodiment, the inter-CM communication unit 41 acquires operation information from other CMs 10, but a primary CM 10 may be provided to centrally manage the operation status. In that case, each CM 10 may be able to grasp the operating status of all CM 10 by communicating only with the primary CM 10 .

In the above-described embodiment, the write-restricted BUD determination unit 42 determined the BUD 20 with the smallest identifier as the normal write BUD 20, but this is not the only option. For example, the write restriction BUD determination unit 4 may determine the normal write BUD 20 in advance. Also, in the above-described embodiment, the write-restricted BUD determination unit 42 determined one BUD 20 to be the normal write BUD 20, but the present invention is not limited to this. For example, a plurality of BUDs 20 may be determined as normal write BUDs 20 .

In the above-described embodiment, the write-restricted BUD determination unit 42 determined the BUD 20 other than the normal write BUD 20 as the write-restricted BUD 20, but the present invention is not limited to this. Also, the write restriction BUD determination unit 42 may determine the write restriction BUD 20 in advance.

In the above-described embodiment, when there are a plurality of write-restricted BUDs 20, the write execution unit 47 similarly writes to each of these write-restricted BUDs 20, but the present invention is not limited to this. For example, the write execution unit 47 may write to these plurality of write-restricted BUDs 20 under different write restrictions.

[4] Supplementary Notes Regarding the above embodiment, the following supplementary notes are disclosed.

(Appendix 1)
A memory control device that controls writing of data to be backed up in a storage control device using a plurality of memory devices having a limited number of writable times,
A memory control device, comprising: a control unit that restricts writing to a part of a plurality of memories based on an operating state of the memory and importance of the data to be backed up.

(Appendix 2)
The control unit writes an amount of the backup data to be written to the part of the memory and an amount of the backup data to be written to the memory other than the part of the memories among the plurality of memories so as to be different. to restrict,
The memory control device according to Supplementary Note 1, characterized by:

(Appendix 3)
The control unit backs up the data to be backed up stored in a volatile memory of the memory control device to the memory.
The memory control device according to appendix 1 or 2, characterized by:

(Appendix 4)
The device according to any one of appendices 1 to 3, further comprising: a state determination unit that determines the operating state of the memory based on at least an operating time of the memory and an accumulated write amount of the memory. memory controller.

(Appendix 5)
and an importance determination unit that determines the importance of the backup target data based on at least a write destination address of the backup target data in a storage area of the memory and a type of the backup target data. 5. The memory control device according to any one of claims 1 to 4.

(Appendix 6)
to the processor of the memory controller,
A memory control device that controls writing of data to be backed up in a storage control device by using a plurality of memory devices with a limited number of writable times, based on the operating state of the memory and the importance of the data to be backed up, Execute processing to restrict writing to some of the multiple memories,
A memory control program characterized by:

(Appendix 7)
a process of restricting writing such that an amount of the backup data to be written to the part of the memory and an amount of the backup data to be written to the memory other than the part of the memories among the plurality of memories are different; causing the processor to execute;
The memory control program according to appendix 6, characterized by:

(Appendix 8)
causing the processor to execute a process of backing up the data to be backed up stored in the volatile memory of the memory control device to the memory;
A memory control program according to appendix 6 or 7, characterized by:

(Appendix 9)
causing the processor to execute a process of determining an operating state of the memory based on at least an operating time of the memory and an accumulated write amount of the memory;
The memory control program according to any one of Appendices 6 to 8, characterized by:

(Appendix 10)
causing the processor to execute a process of determining the degree of importance of the backup target data based on at least a write destination address of the backup target data in a storage area of the memory and a type of the backup target data;
The memory control program according to any one of Appendices 6 to 9, characterized in that:

(Appendix 11)
A memory control device that controls writing of data to be backed up in a storage control device by using a plurality of memory devices with a limited number of writable times, based on the operating state of the memory and the importance of the data to be backed up, a process of restricting writing to a part of memories among a plurality of memories;
A memory control method characterized by:

(Appendix 12)
a process of restricting writing such that the amount of the backup data to be written to the part of the memory and the amount of the backup data to be written to the memory other than the part of the memories among the plurality of memories are different. prepare
The memory control method according to appendix 11, characterized by:

(Appendix 13)
The memory control device comprises a process of backing up the data to be backed up stored in a volatile memory of the memory control device to the memory.
13. The memory control method according to appendix 11 or 12, characterized by:

(Appendix 14)
The memory control device comprises at least a process of determining the operating state of the memory based on the operating time of the memory and an accumulated amount of writing of the memory.
14. The memory control method according to any one of appendices 11 to 13, characterized by:

(Appendix 15)
The memory control device comprises at least a process of determining the importance of the backup target data based on a write destination address of the backup target data in a storage area of the memory and a type of the backup target data.
15. The memory control method according to any one of appendices 11 to 14, characterized by:

1 storage system 10 CM (memory controller)
20 BUD (memory device with limited number of writes)
30 server device 11 CPU
12 storage unit 13 memory (volatile memory)
14 IF unit 15 input/output unit 16 display unit 41 inter-CM communication unit 42 write restriction BUD determination unit 43 importance determination unit 44 status acquisition unit 45 write mode determination unit 46 state determination unit 47 write execution unit (control unit)
50 HDDs
60 Inter-CM communication switch 70 BUD status determination information 71 Importance determination information 72 Write control information 90 Control program 201 Firmware storage area 202 Log information write area 203 Backup area 204 Free area

Claims (9)

A memory control device that controls writing of data to be backed up in a storage control device using a plurality of semiconductor memory devices having a limited number of writable times,
when a plurality of semiconductor memory devices are in operation, determining a portion of the semiconductor memory devices subject to write restriction from among the plurality of semiconductor memory devices based on the identifier of the semiconductor memory device ;
By imposing a write restriction on some of the semiconductor memory devices subject to the restriction based on the life-related status of the semiconductor memory devices and the importance of the data to be backed up, a control unit that delays the time until a semiconductor memory device reaches a failure state, compared to other semiconductor memory devices other than the restricted limited semiconductor memory device among the plurality of semiconductor memory devices. A memory controller, characterized by: The control unit
determining a criticality to be subject to write restriction based on the life-related state of the semiconductor memory device ;
for the data to be backed up with the determined importance level, write restriction is imposed on the part of the semiconductor memory devices subject to the restriction;
2. The memory control device according to claim 1, wherein: The state related to the life of the semiconductor memory device includes a state in which the cumulative write amount of the semiconductor memory device is less than a first threshold, a state in which it is equal to or greater than a second threshold, and a state in which it is greater than or equal to the first threshold and less than a third threshold ( However, the third threshold is less than the second threshold), and the first threshold and the third threshold change according to the operating time of the semiconductor memory device .
3. The memory control device according to claim 1, wherein: The controller controls the amount of the data to be backed up to be written to the limited semiconductor memory devices so that the amount of the data to be backed up to be written to the other semiconductor memory devices is different from the amount of the data to be backed up. write restrictions,
4. The memory control device according to any one of claims 1 to 3, characterized by: The control unit backs up the backup target data stored in the volatile memory of the memory control device to the semiconductor memory device .
The memory control device according to any one of claims 1 to 4, characterized by: A state determination unit that determines a state related to the life of the semiconductor memory device based on at least an operating time of the semiconductor memory device and an accumulated write amount of the semiconductor memory device . 6. The memory control device according to any one of 5. an importance determination unit that determines the importance of the backup target data based on at least a write destination address of the backup target data in a storage area of the semiconductor memory device and a type of the backup target data. The memory control device according to any one of claims 1 to 6, wherein: to the processor of the memory controller,
When the memory control device that performs control to write the data to be backed up in the storage control device using a plurality of semiconductor memory devices with a limited number of writable times is in operation, the plurality of semiconductor memory devices are in operation. out of the semiconductor memory devices , a part of the semiconductor memory devices subject to write restriction is determined based on the identifier of the semiconductor memory device , and the state regarding the life of the semiconductor memory device and the importance of the data to be backed up are determined. Based on this, by imposing a write restriction on some of the semiconductor memory devices subject to the restriction, the time until the semiconductor memory devices subject to the restriction reach a failure state is reduced to the plurality of semiconductor memories. making it slower than other semiconductor memory devices other than some of the semiconductor memory devices to be restricted among devices, causing processing to be executed,
A memory control program characterized by: If a memory control device that controls writing of data to be backed up in a storage control device using a plurality of semiconductor memory devices with a limited number of writable times is in operation, the plurality of semiconductor memory devices A part of the semiconductor memory devices subject to write restriction is determined from among the semiconductor memory devices based on the identifier of the semiconductor memory device , and based on the state regarding the life of the semiconductor memory device and the importance of the data to be backed up. and by imposing a write restriction on the semiconductor memory devices subject to the restriction, the time required for the semiconductor memory devices subject to the restriction to reach a failure state is reduced to the plurality of semiconductor memory devices. slower than other semiconductor memory devices other than the limited semiconductor memory devices,
A memory control method characterized by:

Images