JP2014182674A - Storage device and storage method - Google Patents

Abstract

There is provided a storage device capable of reducing introduction cost and operation cost when a plurality of storage elements having different characteristics are used together as a secondary cache memory.
Data reading and data writing in response to a request from a host terminal, data storage from a main storage means 3 to a primary cache memory 4, and primary cache memory 4 to first secondary cache memory 5 or When the control means 2 for controlling the transfer of data to the second secondary cache memory 6 is provided, the control means 2 may be reused for the data for which the transfer control is performed, and the update frequency is equal to or higher than a predetermined value Transfers data to the first secondary cache memory 5 and transfers the data to the second secondary cache memory 6 when the data can be reused and the update frequency is lower than a predetermined value.
[Selection] Figure 13

Description

  The present invention relates to a storage device having a cache function and a storage method.

  There is a storage system that includes a logical disk that logically divides a storage area of a disk pool and that is used from a host computer connected via a network (SAN: Storage Area Network). The disk pool is composed of a set of magnetic disks such as HDDs (Hard Disk Drives) as in a disk array device represented by RAID (Redundant Array of Inexpensive Disk Drives).

  When RAID is used for the disk pool, the storage system is equipped with a controller that controls RAID. The storage system may also be equipped with a cache memory for caching read / write access to the logical disk from the host computer.

  An SSD (Solid State Drive) is a drive device that uses a flash memory as a storage medium. The SSD has higher performance than the HDD in terms of data transfer speed, and is used as a permanent data storage medium.

  In addition, data to be deleted from the primary cache memory is temporarily stored in the SSD mounted in the disk array device, and when there is an access request again, it is temporarily read from the high-speed SSD instead of being read from the low-speed HDD. There is an SSD secondary cache memory technology that reads data stored in the. By utilizing the high-speed reading technology of SSD, the SSD secondary cache memory technology improves access performance.

  There is an upper limit (lifetime) for the number of SSD writes. When the number of writes reaches the upper limit, the SSD becomes unreadable / writable. When reading / writing becomes impossible, the user is required to replace the target SSD.

  Moreover, due to the difference in recording methods, SSD has a high speed and long life but is expensive SLC (Single Level Cell), and MLC (Multi Level Cell) which is high speed and short life but relatively inexpensive. Exists.

  When an SSD is used as a secondary cache memory, the number of SSD writes tends to increase because rewriting is performed more frequently than when used as a normal storage unit. That is, the SSD used as the secondary cache memory tends to reach the end of its life quickly.

  When the short-life MLC is adopted as the secondary cache memory, the MLC is likely to reach the lifetime if rewriting for storing data to be deleted from the primary cache memory is frequently performed. As a result, medium replacement frequently occurs and operation costs tend to increase.

  When the long-life SLC is adopted as the secondary cache memory, the SLC is less likely to reach the life as compared with the MLC even if rewriting for storing data to be deleted from the primary cache memory is frequently performed. As a result, the operation cost associated with medium replacement can be suppressed. However, since the SLC is an expensive medium, the introduction cost when it is adopted becomes large.

  For example, if the operation is performed using 20 MLCs having an introduction cost of 1 and a lifetime of 1 year, the introduction cost is only 20. However, since all MLCs are exchanged every year, the total cost is 100 when operated for 5 years.

  Further, for example, when an operation is performed using 20 SLC units having an introduction cost of 5 and a lifetime of 10 years, the introduction cost is 100. However, since the SLC is not exchanged for 10 years after the introduction, the total cost remains 100 even after 5 years of operation.

  Therefore, when using SSDs with different characteristics, such as SLC and MLC, as the secondary cache memory, it is possible to improve the access performance of the SSD secondary cache memory while reducing the introduction cost and operation cost of the SSD secondary cache memory. Reduction is a challenge.

  Japanese Patent Application Laid-Open No. 2004-228688 confirms the writing frequency and the number of times of writing to an SSD block in a configuration in which SLC and MLC which are SSDs having different characteristics are mixedly mounted, and in the replacement processing of a block exceeding the upper limit of the number of times of writing It describes a method for reducing costs by allocating long-life SLC blocks to high-blocks so that the allocated blocks are not exhausted and extending their lifetimes.

JP 2010-108246 A

  However, the method described in Patent Document 1 is a method related to an alternative process of a storage unit that stores data in an SSD, and is not assumed when the SSD is used as a secondary cache memory. Therefore, the method described in Patent Document 1 cannot solve the problem when a plurality of SSDs having different characteristics are used in combination as a secondary cache memory.

  Therefore, the present invention provides a storage device and a storage method capable of reducing the introduction cost and the operation cost when a plurality of storage elements having different characteristics are used together as a secondary cache memory.

  A storage device according to the present invention includes a main storage means for storing data, a primary cache memory for caching data stored in the main storage means, and a first secondary for storing data to be deleted from the primary cache memory A cache memory, and a second secondary cache memory that stores data to be deleted from the primary cache memory, using a semiconductor storage medium in which the number of times data can be written is less than the number of times that the first secondary cache memory can be written , Reading and writing data in response to a request from the host terminal, storing data from the main storage means to the primary cache memory, and from the primary cache memory to the first secondary cache memory or the second secondary cache memory Control means for controlling the transfer of data to the control means, the control means If the data to be controlled is reusable and the update frequency is greater than or equal to a predetermined value, the data is transferred to the first secondary cache memory, the data may be reusable, and the update frequency is predetermined If the value is lower than the value, the data is transferred to the second secondary cache memory.

  The storage method according to the present invention comprises a main storage means for storing data, a primary cache memory for caching data stored in the main storage means, and a first secondary for storing data to be deleted from the primary cache memory. A cache memory, and a second secondary cache memory that stores data to be deleted from the primary cache memory, using a semiconductor storage medium in which the number of times data can be written is less than the number of times that the first secondary cache memory can be written A storage method executed in a storage device comprising: reading data and writing data in response to a request from a host terminal; storing data from a main storage means into a primary cache memory; Data to one secondary cache memory or second secondary cache memory If transfer control is performed, the data to be transferred may be reused, and if the update frequency is greater than or equal to a predetermined value, the data is transferred to the first secondary cache memory and may be reused. In addition, when the update frequency is lower than a predetermined value, the data is transferred to the second secondary cache memory.

  According to the present invention, it is possible to reduce the introduction cost and the operation cost when a plurality of storage elements having different characteristics are used together as a secondary cache memory.

1 is a system configuration diagram illustrating a configuration example of an embodiment of a storage device according to the present invention. 2 is a block diagram showing a configuration example of a disk array device 200. FIG. 6 is an explanatory diagram showing an example of access history information stored in a disk array device data storage unit 215. FIG. 6 is an explanatory diagram showing an example of cache allocation management information stored in a disk array device data storage unit 215. FIG. 5 is an explanatory diagram showing an example of primary cache memory management information stored in a disk array device data storage unit 215. FIG. 6 is an explanatory diagram showing an example of secondary cache memory management information stored in a disk array device data storage unit 215. FIG. 6 is an explanatory diagram showing an example of secondary cache memory management information stored in a disk array device data storage unit 215. FIG. It is explanatory drawing which shows an example of the SSD rewrite management information preserve | saved at the disk array apparatus data storage means 215. 4 is a flowchart showing the operation of access history recording means 210. 4 is a flowchart showing the operation of the secondary cache memory storage means 212. It is a flowchart which shows operation | movement of the secondary cache memory resetting means 213. 4 is a flowchart showing the operation of the write upper limit monitoring means 214. It is a block diagram which shows the outline | summary of a memory | storage device.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a system configuration diagram showing a configuration example of an embodiment of a storage device according to the present invention. A disk array device 200 that is a storage device shown in FIG. 1 is connected to a host computer 100 via a communication path 300.

  The host computer 100 issues a data read request or a data write request to the disk array device 200 via the communication path 300.

  The disk array device 200 includes a plurality of secondary cache memories including a first secondary cache memory composed of MLC and a second secondary cache memory composed of SLC. The disk array device 200 records access information in block units of logical disks built in the device.

  When data is stored in the secondary cache memory, the disk array device 200 analyzes the past access tendency with respect to the block of the logical disk in which the data is stored. Then, the disk array device 200 selects the first secondary cache memory or the second secondary cache memory as a data storage destination for the purpose of extending the life of the SSD as much as possible. The disk array device 200 stores the data in the selected secondary cache memory.

  As a specific storage method, the disk array device 200 reads data stored in a block that is determined not to be frequently rewritten and reused in the long term from the past access tendency from the MLC having low write endurance. Store preferentially in the configured first secondary cache memory. In addition, the disk array device 200 stores the data stored in the block that has been determined to be reused but may be updated in the short term, as a second secondary cache composed of SLC having high write resistance. Store in memory.

  In addition, the disk array device 200 is composed of MLC with low write endurance with respect to data continuously stored in the second secondary cache memory composed of SLC with high write endurance without being updated for a long time. To use the first secondary cache memory.

  With the above method, the performance of the SSD secondary cache memory is improved, and the introduction cost and operation cost of the SSD secondary cache memory are reduced.

  The communication path 300 connects the host computer 100 and the disk array device 200 so that they can communicate with each other. The communication path 300 is, for example, a SAN.

  Next, a detailed configuration of the disk array device 200 will be described with reference to FIG. FIG. 2 is a block diagram illustrating a configuration example of the disk array device 200.

  2 includes a storage controller 201, HDDs (Hard Disk Drives) 202 to 204, a pool 205, a logical disk 206, a primary cache memory 207, a secondary cache memory 208, and a secondary cache memory. Cache memory 209, access history recording means 210, access trend analysis means 211, secondary cache memory storage means 212, secondary cache memory resetting means 213, write upper limit monitoring means 214, disk array device data storage Means 215.

  The storage controller 201 has a function of controlling data read processing or write processing based on a read request or write request from the host computer 100. The storage controller 201 stores data from the logical disk 206 to the primary cache memory 207 and stores or discards data stored in the primary cache memory 207 in the secondary cache memory based on instructions from each means.

  The HDDs 202 to 204 are magnetic disks mounted on the disk array device 200. The HDDs 202 to 204 have a function of storing data. The number of magnetic disks mounted on the disk array device 200 may be three or more.

  The pool 205 is a virtual data storage area configured from a set of HDDs 202 to 204. The HDDs 202 to 204 may configure the pool 205 by RAID.

  The logical disk 206 is a virtual data storage area cut out from the pool 205. The storage controller 201 performs data read processing and data write processing on the logical disk 206 instead of the HDD. Two or more logical disks may be cut out from the pool 205.

  The primary cache memory 207 has a function of storing data received by the storage controller 201 in response to a write request from the host computer 100. The primary cache memory 207 has a function of storing data that the storage controller 201 returns to the host computer 100 in response to a read request from the host computer 100.

  The primary cache memory 207 stores data for the most recent read request or write request. However, when it is required to secure a new storage block for storing new data, data that has not been used for a predetermined period is deleted from the primary cache memory 207. The data to be deleted is selected by an algorithm such as LRU (Least Recently Used).

  The secondary cache memory 208 is a first secondary cache memory that stores data to be deleted from the primary cache memory 207. The secondary cache memory 208 is composed of an MLC that is a short-lived SSD.

  The secondary cache memory 209 is a second secondary cache memory that stores data to be deleted from the primary cache memory 207. The secondary cache memory 209 is composed of SLC, which is a long-lifetime SSD.

  The access history recording unit 210 includes, for each address management (block) of the logical disk 206, the number of reads within a predetermined period, the number of writes within a predetermined period, the number of cache hits for the primary cache memory and the secondary cache memory, the update time, It has a function of recording an update interval from the previous time as access history information.

  The access history recording unit 210 operates when the storage controller 201 receives a read request or write request for the logical disk 206 from the host computer 100 or when the storage controller 201 deletes data from the primary cache memory 207.

  The access trend analysis unit 211 has a function of analyzing the access trend for the block of the host computer 100 for each block of the logical disk 206 from the access history information, and deriving the access trend information.

  The secondary cache memory storage unit 212 has a function of determining a candidate for a secondary cache memory that is a storage destination of data to be deleted when data is deleted from the primary cache memory 207.

  The secondary cache memory storage unit 212 selects the first secondary cache memory or the second secondary cache memory as the data storage destination according to the access trend information derived by the access trend analysis unit 211. Next, the secondary cache memory storage unit 212 instructs the storage controller 201 to store data in the selected secondary cache memory.

  Specifically, the secondary cache memory storage unit 212 refers to access tendency information for a block of the logical disk 206 in which data to be deleted from the primary cache memory 207 is stored. The secondary cache memory storage unit 212 confirms the possibility of data reusability and the high update frequency based on the access trend information, and determines the secondary cache memory to be the data storage destination.

  When the data is reusable, the update frequency is low, and the secondary cache memory 208 configured by MLC is available, the secondary cache memory storage unit 212 sets the data storage destination to the secondary cache memory 208. To decide.

  Further, when the data is reusable, the update frequency is high, and the secondary cache memory 209 configured by SLC is available, the secondary cache memory storage unit 212 sets the data storage destination to the secondary cache. The memory 209 is determined.

  If it is determined that the data is not reusable, the secondary cache memory storage unit 212 determines to discard the data without storing it in any of the secondary cache memories.

  The secondary cache memory resetting means 213 changes the storage location of the target data from the second secondary cache memory to the second secondary cache memory when there is data stored beyond the predetermined holding interval in the second secondary cache memory. It has a function of changing to one secondary cache memory.

  Specifically, the secondary cache memory resetting unit 213 confirms whether there is data that has been retained for a long time among the data stored in the secondary cache memory 209 composed of SLC. To do.

  When there is data held beyond a predetermined period and the secondary cache memory 208 configured by MLC is available, the secondary cache memory resetting unit 213 stores the target data in the secondary cache memory 208. Is stored in the storage controller 201.

  Next, the secondary cache memory resetting unit 213 instructs the storage controller 201 to discard the target data stored in the secondary cache memory 209 composed of SLC. In this way, the secondary cache memory resetting unit 213 resets the storage location of data stored in the secondary cache memory 209 beyond a predetermined holding interval.

  The write upper limit monitoring unit 214 has a function of monitoring the number of data writes to the SSD constituting the secondary cache memory. When the SSD write count exceeds a predetermined write upper limit count, the write upper limit monitoring unit 214 limits the use of the SSD exceeding the write upper limit count.

  Specifically, the write upper limit monitoring unit 214 monitors the write process for the SSD constituting the secondary cache memory. When the write count reaches a predetermined threshold value, the write upper limit monitoring unit 214 sets the reached SSD state as a caution state.

  Further, when the write count reaches the write upper limit count, the write upper limit monitoring unit 214 sets the reached SSD state to the unusable state. The write upper limit monitoring unit 214 prevents the target SSD from being used by making it unavailable.

  The disk array device data storage means 215 has a function of storing access history information, cache allocation management information, primary cache memory management information, secondary cache memory management information, and SSD rewrite management information.

  FIG. 3 is an explanatory diagram showing an example of access history information stored in the disk array device data storage means 215. The access history information is read or written by the storage controller 201 in response to a request from the host computer 100, deleted from the primary cache memory 207, and resetted from the secondary cache memory as a data storage destination. Will be updated when

  The access history information stored in the disk array device data storage means 215 includes an LD (logical disk number) 301, an LD address 302, a cache hit count 303, an update time 304, an update interval 305, a read count 306, a write count 307, and It consists of cache update count 308.

  The LD 301 is information indicating the identification number of the logical disk 206 for which the access history is recorded. The LD address 302 is information indicating the block identification number of the logical disk 206 for which the access history is recorded.

  The number of cache hits 303 is the number of times that data exists in the primary cache memory or the secondary cache memory when the host computer 100 makes an access request to the data stored in the block of the target logical disk 206. It is information to show.

  The update time 304 is information indicating the latest time when a write process was performed on a block of the target logical disk 206 in response to a write request from the host computer 100. The update interval 305 is information indicating a difference between the previous update time 304 and the latest update time 304.

  The read count 306 is information indicating the number of times that the read processing for the block of the target logical disk 206 has been performed from the host computer 100. The write count 307 is information indicating the number of times the host computer 100 has performed write processing on the target logical disk 206 block.

  The cache update count 308 is information indicating the number of times data stored in the block of the target logical disk 206 is stored in the primary cache memory or the secondary cache memory.

  FIG. 4 is an explanatory diagram showing an example of cache allocation management information stored in the disk array device data storage means 215. The cache allocation management information is information indicating a storage destination when data of each block of the logical disk 206 is stored in the primary cache memory or the secondary cache memory.

  The cache allocation management information stored in the disk array device data storage unit 215 includes an LD 301, an LD address 302, a cache flag 309, a storage destination 310, and a storage destination address 311.

  The cache flag 309 is information indicating whether the data stored in the block of the target logical disk 206 is stored in the cache memory. The storage destination 310 is information indicating the stored cache memory when data is stored in the cache memory. The storage destination address 311 is information indicating a block of the stored cache memory when data is stored in the cache memory.

  FIG. 5 is an explanatory diagram showing an example of primary cache memory management information stored in the disk array device data storage means 215. The primary cache memory management information is information indicating the data storage status of the primary cache memory 207.

  The primary cache memory management information stored in the disk array device data storage means 215 includes an address 312, a cache flag 313, an LD 301, and an LD address 302.

  The address 312 is information indicating the identification number of the block of the cache memory. The cache flag 313 is information indicating whether or not the target block stores data. An LD 301 and an LD address 302 indicate a logical disk 206 and a block in which stored data is stored.

  FIG. 6 is an explanatory diagram showing an example of secondary cache memory management information stored in the disk array device data storage means 215. The secondary cache memory management information is information indicating the data storage status of the secondary cache memory 208 and the secondary cache memory 209. FIG. 7 is an explanatory diagram showing another example of the secondary cache memory management information stored in the disk array device data storage unit 215.

  The secondary cache memory management information stored in the disk array device data storage unit 215 includes an SSD 314, an address 315, a cache flag 316, an LD 301, an LD address 302, and a holding time 317.

  The SSD 314 is information indicating the identification number of the SSD constituting the target secondary cache memory. The address 315 is information indicating the identification number of the target SSD block. The cache flag 316 is information indicating whether or not the target block stores data.

  An LD 301 and an LD address 302 indicate a logical disk 206 and a block in which stored data is stored. The retention time 317 is information indicating the time when data is stored when the target block stores data.

  FIG. 8 is an explanatory diagram showing an example of SSD rewrite management information stored in the disk array device data storage means 215. The SSD rewrite management information is information indicating the rewrite status of the SSD. The SSD rewrite management information holds information for determining whether the target SSD can be used.

  The SSD rewrite management information stored in the disk array device data storage unit 215 includes an SSD 314, a type 318, a write count 319, an upper limit write count 320, a threshold value 321 and a state 322.

  The type 318 is information indicating the type of SSD corresponding to the SSD 314. The write count 319 is information indicating the sum of the write counts in each block of the target SSD.

  The upper limit number of times of writing 320 is information indicating the number of times of writing at which writing to the target SSD becomes impossible. The threshold value 321 is information indicating the number of times the target SSD is in a “Caution” state.

  The state 322 is information indicating the current state of the target SSD. In the status 322, one of “normal”, “caution”, and “unusable” is displayed. When the state 322 is “normal”, the target SSD can be written normally.

  When the state 322 is “Caution”, the target SSD is in a state before writing becomes impossible. When the status 322 is “unusable”, the target SSD cannot be written.

  The storage controller 201, the access history recording unit 210, the access tendency analysis unit 211, the secondary cache memory storage unit 212, the secondary cache memory resetting unit 213, and the write upper limit monitoring unit 214 in this embodiment are, for example, a program This is realized by a CPU (Central Processing Unit) that operates according to the above.

  The primary cache memory 207 is realized by, for example, an SRAM (Static Random Access Memory) that can be directly read and written by the CPU.

  The operation of the present invention will be described below. First, an operation of update processing of the access history recording unit 210 when the host computer 100 issues a read request or a write request, or when the storage controller 201 performs a deletion process of data stored in the primary cache memory 207. Will be described with reference to the flowchart of FIG. FIG. 9 is a flowchart showing the operation of the access history recording unit 210.

  The access history recording unit 210 acquires the LD 301 and the LD address 302 indicating the block of the logical disk 206 that is the target of the access request. At the same time, the access history recording unit 210 acquires current time information (step S101).

  Next, the access history recording unit 210 confirms whether or not the access request is a request to delete data in the primary cache memory 207 from the storage controller 201 (step S102). If the access request is not a data deletion request (No in step S102), the access history recording unit 210 advances the process to step S105.

  When the access request is a deletion request for data in the primary cache memory 207 by the storage controller 201 (Yes in step S102), the access history recording unit 210 performs the storage process in the secondary cache memory that is executed in accordance with the deletion request. Whether the data is stored in the secondary cache memory is checked (step S103). When the data is not stored (No in step S103), the access history recording unit 210 ends the process.

  When data is stored in the secondary cache memory (Yes in step S103), the access history recording unit 210 includes the LD 301, LD of the secondary cache memory management information corresponding to the block of the secondary cache memory in which the data is stored. The address 302 and the holding time 317 are each updated (step S104).

  Further, the access history recording unit 210 updates the cache update count 308 of the access history information corresponding to the block of the logical disk 206 in which the stored data is stored (step S105).

  Further, the access history recording unit 210 updates the write count 319 of the SSD rewrite management information corresponding to the SSD constituting the secondary cache memory in which the data is stored (step S106). After the update, the access history recording unit 210 ends the process.

  In step S105, the access history recording unit 210 confirms whether the access request is a write request by the host computer 100 (step S107).

  When the access request is a write request by the host computer 100 (Yes in step S107), the access history recording unit 210 sets the write history information write count 307 and the update time 304 corresponding to the address of the logical disk 206 to be written. Each is updated (step S108).

  At the same time, the access history recording unit 210 calculates the difference between the update time 304 of the previous write request and the current time, and updates the update interval 305 (step S108). After the update, the access history recording unit 210 ends the process.

  When the access request is not a write request by the host computer 100, that is, when it is a read request (No in step S107), the access history recording unit 210 confirms whether or not the target data for the read request exists in the cache memory. (Step S109). If it does not exist in the cache memory (No in step S109), the access history recording unit 210 advances the process to step S111.

  If it exists in the cache memory (Yes in step S109), the access history recording unit 210 determines that the data targeted for the read request is reused from the cache memory. Next, the access history recording unit 210 updates the cache hit count 303 of the access history information corresponding to the block of the logical disk 206 in which the target data is stored (step S110).

  Next, the access history recording unit 210 updates the read count 306 of the access history information corresponding to the block of the logical disk 206 in which the data to be read is stored (step S111). After the update, the access history recording unit 210 ends the process.

  Next, the operation of the storage process of the secondary cache memory storage unit 212 when the process of deleting the data stored in the primary cache memory 207 is performed will be described with reference to the flowchart of FIG. FIG. 10 is a flowchart showing the operation of the secondary cache memory storage unit 212.

  The access trend analysis unit 211 refers to the access history information and analyzes the past access trend with respect to the block of the logical disk 206 in which the data to be deleted is stored (step S201). Specifically, the access trend analysis unit 211 refers to the access history information and calculates an average update interval, an average cache hit count, and a read rate.

  The secondary cache memory storage unit 212 refers to the access trend information from the access trend analysis unit 211. Then, the secondary cache memory storage unit 212 confirms whether the average cache hit count of the block of the target logical disk 206 exceeds a predetermined value or whether the read rate exceeds a predetermined value (step S202).

  If the average cache hit count is less than or equal to the predetermined value and the read rate is also less than or equal to the predetermined value (No in step S202), the secondary cache memory storage unit 212 re-stores the data stored in the block of the target logical disk 206. Judge that there is no availability.

  The secondary cache memory storage unit 212 instructs the storage controller 201 to discard the target data stored in the primary cache memory 207. Receiving the instruction, the storage controller 201 discards the target data from the primary cache memory 207 (step S206).

  Next, since the data is discarded, the secondary cache memory storage unit 212 stores the cache allocation management information corresponding to the block of the logical disk 206 in which the target data is stored, and the primary cache memory in which the target data is stored. The primary cache memory management information corresponding to the block 207 is updated to a state in which no data is stored (step S208). After the update, the secondary cache memory storage unit 212 ends the process.

  When the average number of cache hits exceeds a predetermined value, or when the read rate exceeds a predetermined value (Yes in step S202), the secondary cache memory storage unit 212 is stored in a block of the target logical disk 206. It is determined that the data that can be reused is reusable.

  Next, the secondary cache memory storage unit 212 confirms whether the average update interval for the block of the logical disk 206 in which the data to be deleted is stored is greater than or equal to a predetermined value and the cache update count 308 is less than or equal to the predetermined value. (Step S203).

  When the average update interval is not less than the predetermined value and the cache update count 308 is not more than the predetermined value (Yes in step S203), the secondary cache memory storage unit 212 stores the data stored in the block of the target logical disk 206. However, it is determined that the data is reused but the update frequency is low. That is, the secondary cache memory storage unit 212 determines that the data to be deleted can be stored in the secondary cache memory 208 composed of MLC with low write endurance.

  The secondary cache memory storage unit 212 refers to the SSD rewrite management information to check whether the secondary cache memory 208 is usable. Even when the state of the SSD constituting the secondary cache memory 208 is “notice” when referring to the SSD rewrite management information, the secondary cache memory storage unit 212 also stores the secondary cache memory. 208 may be determined as unusable.

  If it is determined in step S203 that the secondary cache memory 208 can be stored, and the secondary cache memory 208 is available, the secondary cache memory storage unit 212 stores the target data stored in the primary cache memory 207 as a second data. The storage controller 201 is instructed to store in the next cache memory 208. Receiving the instruction, the storage controller 201 stores the target data in the secondary cache memory 208 (step S204).

  If it is determined in step S203 that the secondary cache memory 208 can be stored, but the secondary cache memory 208 is unavailable, the secondary cache memory storage unit 212 stores the target data in the secondary cache memory if available. 209 may be stored. When the secondary cache memory 209 is also unavailable, the secondary cache memory storage unit 212 may discard the target data.

  When the average update interval is less than the predetermined value, or when the cache update count 308 is larger than the predetermined value (No in step S203), the secondary cache memory storage unit 212 is stored in the block of the target logical disk 206. It is determined that the existing data is reused and frequently updated. That is, the secondary cache memory storage unit 212 determines that the data to be deleted can be stored in the secondary cache memory 209 composed of an SLC with high write endurance.

  The secondary cache memory storage unit 212 refers to the SSD rewrite management information and checks whether the secondary cache memory 209 is available. Even when the state of the SSD constituting the secondary cache memory 209 is “notice” when referring to the SSD rewrite management information, the secondary cache memory storage unit 212 also stores the secondary cache memory. 209 may be determined as unusable.

  If it is determined in step S203 that the secondary cache memory 209 can be stored and the secondary cache memory 209 is available, the secondary cache memory storage unit 212 stores the target data stored in the primary cache memory 207 as the secondary cache memory 209. Instructs the storage controller 201 to store in the next cache memory 209. Receiving the instruction, the storage controller 201 stores the target data in the secondary cache memory 209 (step S205).

  If it is determined in step S203 that the secondary cache memory 209 can be stored, but the secondary cache memory 209 is unavailable, the secondary cache memory storage unit 212 stores the target data in the secondary cache memory if available. 208 may be stored. When the secondary cache memory 208 is also unavailable, the secondary cache memory storage unit 212 may discard the target data.

  After the processing in step S204 or step S205, the secondary cache memory storage unit 212 stores the data in the secondary cache memory for the access history information corresponding to the block of the logical disk 206 in which the target data is stored. This is reflected (step S207).

  Next, the secondary cache memory storage unit 212 reflects that the data is stored in the secondary cache memory in the cache allocation management information corresponding to the block of the logical disk 206 in which the target data is stored ( Step S209).

  Further, the secondary cache memory storage unit 212 stores the target data in a state where the data is not stored in the primary cache memory management information corresponding to the block of the primary cache memory 207 in which the target data was stored. The secondary cache memory management information corresponding to the secondary cache memory block is updated to a state where data is stored (step S209).

  In addition, the secondary cache memory storage unit 212 updates the write count 319 of the SSD rewrite management information corresponding to the SSD constituting the secondary cache memory in which the data is stored (step S209). After the update, the secondary cache memory storage unit 212 ends the process.

  Next, the storage processing of the secondary cache memory resetting unit 213 when it is confirmed whether or not data that has not been frequently rewritten is stored in the secondary cache memory 209 composed of SLC having high write endurance. Will be described with reference to the flowchart of FIG. FIG. 11 is a flowchart showing the operation of the secondary cache memory resetting means 213.

  The secondary cache memory resetting means 213 refers to the secondary cache memory management information corresponding to the secondary cache memory 209, and periodically analyzes the cache holding status for each block of the secondary cache memory 209 (step S301). .

  Specifically, the secondary cache memory resetting unit 213 calculates the difference between the retention time 317 of each block of the secondary cache memory 209 and the current time, and calculates a difference time that is a cache retention period.

  The secondary cache memory resetting unit 213 checks whether or not the calculated cache retention period is equal to or greater than a predetermined value (step S302). When the cache retention period is less than the predetermined value (No in step S302), the secondary cache memory resetting unit 213 ends the process for the target block.

  When the cache retention period is equal to or longer than the predetermined value (Yes in step S302), the update frequency of the data stored in the target block is low.

  Therefore, the secondary cache memory resetting unit 213 determines that the target data can be reset in the secondary cache memory 208 configured by MLC having low write endurance. However, when the secondary cache memory 208 is in an unusable state as a result of referring to the SSD rewrite management information, the secondary cache memory resetting unit 213 determines that resetting to the secondary cache memory 208 is impossible.

  If it is determined in step S302 that resetting is possible, the secondary cache memory resetting unit 213 instructs the storage controller 201 to store the target data from the secondary cache memory 209 into the secondary cache memory 208. Receiving the instruction, the storage controller 201 stores the target data in the secondary cache memory 208 (step S303).

  After the storage, the secondary cache memory resetting means 213 receives the access history information and cache allocation management information corresponding to the block of the logical disk 206 in which the target data is stored, and the secondary cache memory in which the target data is stored. The secondary cache memory management information corresponding to the 208 block and the secondary cache memory 209 block is updated (step S304).

  Further, the secondary cache memory resetting unit 213 updates the write count 319 of the SSD rewrite management information corresponding to the SSD constituting the secondary cache memory storing the data (step S304).

  After updating each management information, the secondary cache memory resetting unit 213 instructs the storage controller 201 to discard the target data stored in the secondary cache memory 209. Receiving the instruction, the storage controller 201 discards the target data from the secondary cache memory 209 (step S305). After discarding, the secondary cache memory resetting means 213 terminates the processing.

  Next, when issuing a write request by the host computer 100 or performing storage processing in the secondary cache memory by the storage controller 201, the write when checking the write upper limit of the SSD used as the secondary cache memory The monitoring processing operation of the upper limit monitoring unit 214 will be described with reference to the flowchart of FIG. FIG. 12 is a flowchart showing the operation of the write upper limit monitoring unit 214.

  The write upper limit monitoring unit 214 refers to the SSD rewrite management information, and analyzes the status of the SSD constituting the secondary cache memory to be accessed (step S401). Then, the write upper limit monitoring unit 214 confirms whether the write count 319 of the target SSD (each of SLC and MLC) exceeds the upper limit write count 320 (step S402).

  When the write count 319 exceeds the maximum write count 320 (Yes in step S402), the write upper limit monitoring unit 214 updates the state 322 of the SSD rewrite management information corresponding to the target SSD to “unusable” (step S402). S404). After the update, the write upper limit monitoring unit 214 ends the monitoring process for the target SSD.

  When the write count 319 is less than the upper limit write count 320 (No in step S402), the write upper limit monitoring unit 214 confirms whether the SSD write count 319 exceeds the threshold 321 (step S403). If the write count 319 does not exceed the threshold value 321 (No in step S403), the write upper limit monitoring unit 214 ends the monitoring process for the target SSD.

  When the write count 319 exceeds the threshold value 321 (Yes in step S403), the write upper limit monitoring unit 214 updates the state 322 of the SSD rewrite management information corresponding to the target SSD to “caution” (step S405). ). After the update, the write upper limit monitoring unit 214 ends the monitoring process for the target SSD.

  The storage device of the present embodiment can use a plurality of secondary cache memories composed of SSDs having different characteristics according to the update frequency of data to be stored.

  Specifically, the storage device stores the first secondary cache constituted by the MLC with a low storage cost for the data that is determined to have a high possibility of being reused in the long term without being updated based on the past access tendency. Arrange to use memory. Further, data that is determined to be reused but is updated in the short term is arranged so as to use the second secondary cache memory that is composed of SLC having a long life and high storage cost.

  In addition, the storage device rearranges data stored in the second secondary cache memory configured by SLC for a long time and reused in the first secondary cache memory configured by MLC. By these means, it is possible to reduce the rewriting frequency for the MLC having a short life as much as possible, to prolong the life of the MLC, and to suppress the replacement frequency of the MLC when the life is reached.

  Therefore, a user who uses the storage device of this embodiment as a disk array device equipped with a secondary cache memory composed of a plurality of SSDs (SLC and MLC) having different characteristics and costs has a low-cost SSD life span. It becomes possible to extend the time, and the total cost for the introduction and operation of the SSD secondary cache memory can be reduced.

  For example, if 10 MLCs with an introduction cost of 1 and a lifetime of 1 year and 10 SLCs with an introduction cost of 5 and a lifetime of 10 years are used together for 5 years, the total cost of the technology so far is 50 + 50 100. On the other hand, since the frequency of MLC replacement can be suppressed by using the storage device of the present embodiment, the MLC replacement is not required at the best for five years, and the total cost can be suppressed to 60 of 10 + 50.

  Next, the outline of the present invention will be described. FIG. 13 is a block diagram illustrating an outline of the storage device. The storage device 1 according to the present invention includes a main storage unit 3 (for example, a logical disk 206) for storing data, and a primary cache memory 4 (for example, a primary cache memory 207) for caching data stored in the main storage unit 3. ), The first secondary cache memory 5 (for example, the secondary cache memory 209) that stores data to be deleted from the primary cache memory 4, and the number of times the data can be written is written to the first secondary cache memory 5. A semiconductor storage medium less than the possible number of times is used, a second secondary cache memory 6 (for example, secondary cache memory 208) for storing data to be deleted from the primary cache memory 4, and a request from the host terminal Data read and data write, main memory means 3 to primary cache memory 4 Control means 2 (for example, a storage controller 201) for storing the data and controlling the transfer of data from the primary cache memory 4 to the first secondary cache memory 5 or the second secondary cache memory 6. The control means 2 transfers the data to the first secondary cache memory 5 when there is a possibility of reusing the data for which transfer control is performed and the update frequency is equal to or higher than a predetermined value. If the update frequency is lower than the predetermined value, the data is transferred to the second secondary cache memory 6.

  With such a configuration, when a storage device is used, a user who uses the storage device can reduce the introduction cost and the operation cost when using a plurality of storage elements having different characteristics as a secondary cache memory.

  The storage device 1 includes access information storage means (for example, disk array device data storage means 215) that stores access information including information on the reusability of each data stored in the main storage means 3 and the update frequency. The control unit 2 may perform data transfer control with reference to the access information.

  With such a configuration, when the storage device is used, the user who uses the storage device can save the access information of each data.

  The storage device 1 has a resetting unit (for example, a second resetting unit) that instructs the control unit 2 to move the data stored in the first secondary cache memory 5 for a predetermined period or longer to the second secondary cache memory 6. Next cache memory resetting means 213) may be provided.

  With such a configuration, when the storage device is used, the user who uses the storage device can extend the life of the secondary cache memory including the MLC having low write endurance.

  The storage device 1 checks the number of writes to each of the semiconductor storage medium used as the first secondary cache memory 5 and the semiconductor storage medium used as the second secondary cache memory 6, and the number of writes is There may be provided monitoring means (for example, write upper limit monitoring means 214) that disables the use of the semiconductor storage medium used as the secondary cache memory that exceeds the threshold when the predetermined threshold is exceeded. Good.

  With such a configuration, when the storage device is used, writing to the SSD is stopped before the number of times of writing reaches the upper limit, so that the user who uses the storage device can avoid the state in which the SSD cannot be read and written.

1 storage device 2 control means 3 main storage means 4 primary cache memory 5 first secondary cache memory 6 second secondary cache memory 100 host computer 200 disk array device 201 storage controllers 202 to 204 HDD
205 Pool 206 Logical disk 207 Primary cache memory 208, 209 Secondary cache memory 210 Access history recording means 211 Access trend analysis means 212 Secondary cache memory storage means 213 Secondary cache memory resetting means 214 Write upper limit monitoring means 215 Disk array device Data storage means 300 communication path 301 LD
302 LD address 303 Cache hit count 304 Update time 305 Update interval 306 Read count 307 Write count 308 Cache update count 309 Cache flag 310 Storage destination 311 Storage destination address 312 Address 313 Cache flag 314 SSD
315 Address 316 Cache flag 317 Retention time 318 Type 319 Write count 320 Upper limit write count 321 Threshold 322 Status

Claims (8)

Main storage means for storing data;
A primary cache memory for caching data stored in the main storage means;
A first secondary cache memory for storing data to be deleted from the primary cache memory;
A second secondary cache memory for storing data to be deleted from the primary cache memory, using a semiconductor storage medium in which the number of times data can be written is less than the number of writeable times of the first secondary cache memory;
Reading data and writing data in response to a request from the host terminal, storing data from the main storage means to the primary cache memory, and from the primary cache memory to the first secondary cache memory or the second Control means for controlling the transfer of data to the secondary cache memory,
The control means transfers the data to the first secondary cache memory when the data subjected to the transfer control can be reused and the update frequency is equal to or higher than a predetermined value, and re-transfers the data to the data. The storage device, wherein the data is transferred to the second secondary cache memory when there is a possibility of use and the update frequency is lower than the predetermined value. Access information storage means for storing access information including information on reusability of each data stored in the main storage means and update frequency,
The storage device according to claim 1, wherein the control unit performs data transfer control with reference to the access information. The storage according to claim 1 or 2, further comprising resetting means for instructing the control means to move data stored in the first secondary cache memory for a predetermined period or longer to the second secondary cache memory. apparatus. The number of writes to each of the semiconductor storage medium used as the first secondary cache memory and the semiconductor storage medium used as the second secondary cache memory is confirmed, and the number of writes reaches a predetermined threshold value. 4. The monitoring device according to claim 1, further comprising a monitoring unit that disables use of the semiconductor storage medium that is used as a secondary cache memory that exceeds the threshold when exceeding the threshold value. 5. The storage device described. Main storage means for storing data, a primary cache memory for caching data stored in the main storage means, a first secondary cache memory for storing data to be deleted from the primary cache memory, and data A memory comprising a second secondary cache memory for storing data to be deleted from the primary cache memory, wherein a semiconductor storage medium having a smaller number of writable times than the first secondary cache memory is used. A storage method executed in a device comprising:
Read and write data in response to a request from the host terminal,
Storing data from the main storage means to the primary cache memory;
Performing transfer control of data from the primary cache memory to the first secondary cache memory or the second secondary cache memory;
The data for which the transfer control is performed may be reused, and if the update frequency is equal to or higher than a predetermined value, the data is transferred to the first secondary cache memory, and the data may be reused. And when the update frequency is lower than the predetermined value, the data is transferred to the second secondary cache memory. Storing access information including information on reusability and update frequency of each data stored in the main storage means;
The storage method according to claim 5, wherein data transfer control is performed with reference to the access information. The storage method according to claim 5 or 6, wherein data stored in the first secondary cache memory for a predetermined period or more is moved to the second secondary cache memory. The number of writes to each of the semiconductor storage medium used as the first secondary cache memory and the semiconductor storage medium used as the second secondary cache memory is confirmed, and the number of writes reaches a predetermined threshold value. The storage method according to any one of claims 5 to 7, wherein the semiconductor storage medium that is used as a secondary cache memory that exceeds the threshold when the threshold is exceeded is unavailable. .

Images