[go: up one dir, main page]

WO2018189847A1 - Dispositif de mémorisation et procédé de gestion de mémoire cache - Google Patents

Dispositif de mémorisation et procédé de gestion de mémoire cache Download PDF

Info

Publication number
WO2018189847A1
WO2018189847A1 PCT/JP2017/015040 JP2017015040W WO2018189847A1 WO 2018189847 A1 WO2018189847 A1 WO 2018189847A1 JP 2017015040 W JP2017015040 W JP 2017015040W WO 2018189847 A1 WO2018189847 A1 WO 2018189847A1
Authority
WO
WIPO (PCT)
Prior art keywords
cache
capacity
value
area
partition area
Prior art date
Application number
PCT/JP2017/015040
Other languages
English (en)
Japanese (ja)
Inventor
小百合 松井
純二 岩崎
恭平 神野
英美子 藤松
Original Assignee
株式会社日立製作所
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 株式会社日立製作所 filed Critical 株式会社日立製作所
Priority to PCT/JP2017/015040 priority Critical patent/WO2018189847A1/fr
Publication of WO2018189847A1 publication Critical patent/WO2018189847A1/fr

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems

Definitions

  • the present invention relates to a storage apparatus and a cache memory management method, and is suitable for application to, for example, a storage apparatus and a cache memory management method for controlling an influence range of a load of a cache memory.
  • the read / write data of the specific host device may occupy a large area of the cache memory. is there. Under such circumstances, the other host devices must wait for writing to the cache memory, and there is a possibility that the data writing speed of the other host devices may be reduced.
  • the cache memory built in the storage system is logically divided, and the available cache capacity is allocated to the host device in advance, so that a situation where a specific host device occupies a large area of the cache memory is avoided.
  • a technique for controlling the range of influence of load during use is disclosed (see Patent Document 1).
  • the cache memory of the storage system is logically divided using the above-mentioned technology, it is performed based on customer requirements, future usage plans, etc., but the logically divided cache is due to the fact that the usage status after the start of operation is different from the assumption.
  • a situation occurs in which the load for each memory (hereinafter referred to as a cache partition area) is uneven.
  • the cache partition area is used more than when planned for use and there is a concern about the impact of the system, it is necessary to take measures to add allocated capacity. Since preparations such as procurement of cache memory, design for additional work, and customer adjustment are made, there is a problem that it takes time to deal with it and measures are delayed.
  • the present invention has been made in consideration of the above points, and intends to propose a storage apparatus that can effectively use a cache memory.
  • a cache capacity that can be used in each of a plurality of cache division areas obtained by logically dividing the cache memory is allocated in advance to the host device, and the load on the cache memory due to the input / output request of the host device
  • Compare and redistribute cache capacity A relocation review unit that determines whether or not the configuration is possible, and a configuration change unit that rearranges the cache capacity of the plurality of cache division regions based on a result of the review of the relocation review unit I made it.
  • a cache capacity that can be used in each of a plurality of cache divided areas obtained by logically dividing the cache memory is allocated to the host device in advance, and the influence range of the load on the cache memory due to the input / output request of the host device is controlled.
  • the cache partition area cache is leveled so that the load of the cache partition area is leveled based on the cache write wait information monitored for each of the plurality of cache partition areas obtained by logically partitioning the cache memory. Capacity relocation is performed.
  • the cache memory can be used effectively.
  • reference numeral 1 denotes a computer system according to a first embodiment as a whole.
  • the computer system 1 includes a plurality of servers 2 (in this example, a server 2A, a server 2B, a server 2C,...) And a storage device 3, and the plurality of servers 2 and the storage device 3 have communication paths. Connected through.
  • the communication path may include a communication network such as a SAN (Storage Area Network).
  • SAN Storage Area Network
  • the server 2A includes a storage device such as a CPU (Central Processing Unit), RAM (Random Access Memory), ROM (Read Only Memory), HDD (Hard Disc Drive), one or more NICs (Network Interface Card), etc. It is a host device such as a personal computer, a workstation, or a mainframe.
  • the server 2A includes an FC (Fibre Channel) port 21A provided in a communication interface such as an HBA (Host Bus Adapter), and is communicably connected to the storage apparatus 3 via the FC port 21A.
  • the server 2A provides a service to each client terminal by, for example, reading / writing data from / to the storage apparatus 3 in response to a request from each client terminal (not shown).
  • the storage device 3 includes a storage device 30 and a controller 31 that controls input / output of data to / from the storage device 30.
  • the storage device 30 includes one or a plurality of disks such as a SCSI (Small Computer System Interface) disk, a SATA (Serial AT Attachment) disk, and an optical disk.
  • One or more logical volumes VOL are set on the physical storage area provided by the storage device 30. Data from the server 2 is stored in units of blocks of a predetermined size (hereinafter referred to as logical blocks) in the logical volume VOL.
  • a unique volume number is assigned to each logical volume VOL.
  • data input / output is performed by using the combination of this volume number and the block number (LBA: Logical Block ⁇ Address) of the logical block assigned to each logical block as the address. It is done by specifying.
  • LBA Logical Block ⁇ Address
  • the controller 31 includes one or a plurality of channel adapters 32, a microprocessor 33, a cache memory 34, and a disk I / F 35.
  • the controller 31 receives this I / O. Data is read from and written to the storage device 30 as requested.
  • the channel adapter 32 is a hardware device that performs protocol control during communication with the server 2A via the communication path, and includes a plurality of ports 36 (port 36A, port 36B, and port 36C in this example) connected to the communication path. , Communicate with the server 2 via the port 36.
  • the microprocessor 33 is a semiconductor chip having a memory, a CPU, and the like.
  • the memory holds control information and programs necessary for processing executed by the CPU.
  • the CPU specifies an access target device and processes the I / O request.
  • the CPU specifies a device to be accessed by a LUN (Logical Unit Unit Number) included in the I / O request.
  • various functions of the microprocessor 33 are executed by the CPU executing the program stored in the memory.
  • the cache memory 34 stores data in advance or temporarily stores data received from the server 2 in response to an I / O request from the server 2. This increases the processing speed of access requests.
  • the storage area (cache capacity) of the cache memory 34 is managed as a plurality of logically divided areas (cache divided areas).
  • FCport 21A is associated with port 36A
  • port 36A is associated with a predetermined logical volume VOL
  • the predetermined logical volume VOL is associated with one cache partition area.
  • an available cache capacity is allocated to the server 2 in advance to avoid a situation in which the specific server 2 occupies a large area of the cache memory 34, and the data in the storage device 30 is input. The range of influence of load during output is controlled.
  • the storage apparatus 3 acquires the configuration information and operation status of the cache memory 34, analyzes them, and based on threshold values for determining expansion and contraction of each cache partition area The allocation of cache partition areas is automatically reviewed periodically. As a result, the load on the cache division area is leveled. Details will be described below.
  • FIG. 2 shows an example of a configuration related to the microprocessor 33 of the storage apparatus 3.
  • the microprocessor 33 stores a threshold management program, a monitoring management program, a rearrangement review program, a configuration change program, and a cache memory management program.
  • the threshold management unit 201 that manages the configuration information (threshold, caution value, etc.) of the cache memory 34 is realized.
  • the threshold value is a predetermined value that is used when determining whether or not the cache capacity of each cache partition area is insufficient, when calculating the insufficient cache capacity or the removable cache capacity, and the like. is there.
  • the attention value is a value used in the same manner as the threshold value (threshold value in a broad sense), but is a value lower than the threshold value and defined in advance.
  • the monitoring management unit 202 that manages the operation status of the cache memory 34 is realized.
  • the monitoring management unit 202 acquires cache write wait information for each cache partition area as the operation status.
  • the monitoring management unit 202 monitors the cache memory 34 and acquires monitoring data obtained by monitoring the cache write waiting rate of each cache divided area at predetermined time intervals.
  • the relocation examination unit 203 that examines the allocation (relocation) of the cache capacity of the cache divided area based on the configuration information and the operation status of the cache memory 34 is realized.
  • the relocation examination unit 203 confirms whether or not there is a cache partition area exceeding the threshold (a cache partition area exceeding the threshold).
  • the reallocation review unit 203 calculates a cache capacity that is insufficient in the cache partition area that exceeds the threshold value, and determines a predetermined cache partition area (other than the cache partition area that exceeds the threshold value).
  • the cache capacity that can be deleted is calculated for one or more cache partition areas), and the cache capacity that can be deleted is compared with the cache capacity that can be deleted to determine whether the cache capacity can be relocated. To do.
  • the relocation examination unit 203 checks whether or not there is a cache partition area (a cache partition area exceeding the caution value) that does not exceed the threshold value but exceeds the caution value.
  • the relocation review unit 203 calculates a cache capacity that needs to be added in the cache partition area that exceeds the caution value, and performs cache writing of the cache partition area that exceeds the caution value.
  • the cache capacity that can be deleted is calculated for one or more cache partition areas that have a cache write waiting ratio that is smaller than the waiting ratio, and the cache capacity that needs to be added is compared with the cache capacity that can be deleted. It is determined whether or not rearrangement is possible.
  • the configuration change unit 204 that changes the configuration information of the cache partition area based on the examination result of the relocation examination unit 203 is realized.
  • the cache memory management unit 205 that logically divides and manages the storage area of the cache memory 34 based on the configuration information of the cache division area is realized.
  • the microprocessor 33 includes a management database 206, and various tables (such as the table shown in FIG. 3) are stored in the management database 206.
  • FIG. 3 shows an example of a table stored in the management database 206.
  • the management database 206 includes a threshold management table 301, a monitoring information table 302, a reference value table 303, a cache monitoring data table 304, a threshold exceeding target table 305, and an attention value exceeding target table 306.
  • the rearrangement information table 307 and the storage cache configuration information table 308 are stored. Each table will be described with reference to FIGS.
  • FIG. 4 shows an example of the threshold management table 301.
  • the threshold management table 301 stores information indicating the threshold value of the cache write wait rate and information indicating the attention value of the cache write wait rate.
  • the cache write waiting rate of the cache partition area exceeds the threshold, it indicates that the cache partition area is heavily loaded and the cache capacity is insufficient.
  • the cache write waiting rate of the cache partition area does not exceed the threshold value but exceeds the caution value, it indicates that the load of the cache partition area is appropriate.
  • the cache write waiting rate in the cache partition area is below the caution value, it indicates that the load on the cache partition area is low and the cache capacity is sufficient.
  • the cache performance wait rate “the highest performance limit value / performance attention value” (50% in this example) guided by the product is set.
  • the caution value “the lowest value of the performance caution value” (30% in this example) guided by the product is set. Note that the caution value may be a value exceeding the appropriate performance value guided by the product.
  • the threshold value and the caution value are not limited to the above-described values, and appropriate values can be used.
  • FIG. 5 shows an example of the monitoring information table 302.
  • the monitoring information table 302 stores information on a periodic check interval indicating an interval at which the load of the cache partition area is leveled (cache capacity rearrangement) according to the operation status of the cache partition area. Information on the periodic check interval can be changed by the administrator according to the service level.
  • FIG. 6 shows an example of the reference value table 303.
  • the time (time) when the MAX value of the cache write wait rate is monitored, the MAX value of the cache write wait rate, the number of times the write wait rate exceeds the warning value, and the cache write Information on the number of times the waiting rate threshold has been exceeded is stored in association with each other.
  • the MAX value of the cache write wait rate indicates the maximum value of the cache write wait rate of the cache divided area monitored at the regular check interval.
  • the number of times that the write wait rate caution value is exceeded indicates the number of times that the cache write wait rate of the cache divided area monitored in the regular check interval exceeds the caution value but falls below the threshold value.
  • the number of times the cache write wait rate exceeds the threshold indicates the number of times that the cache write wait rate of the cache divided area monitored in the regular check interval exceeds the threshold.
  • FIG. 7 shows an example of the cache monitoring data table 304.
  • the cache monitoring data table 304 is provided for each cache partition area.
  • FIG. 7 shows an example of the first cache partition area. Note that the configuration is not limited to the configuration in which the cache monitoring data table 304 is provided for each cache division area, and a configuration in which the management is collectively performed by one table may be employed.
  • the cache monitoring data table 304 information on the time (time) when the cache write wait rate of the cache divided area is monitored and the cache write wait rate at that time is stored in association with each other.
  • the monitoring interval is set every predetermined time (5 minutes in this example).
  • the predetermined time can be changed by the administrator.
  • FIG. 8 shows an example of the threshold value exceeding target table 305.
  • the threshold value exceeding target table 305 stores information on the threshold exceeding cache division region (target part cache), the MAX value of the cache write waiting rate, and the insufficient capacity in association with each other.
  • This insufficient capacity is a cache capacity that is insufficient in the cache division area that exceeds the threshold, and is a cache capacity that is necessary for the MAX value of the cache write wait rate to be a predetermined cache capacity that is at least the threshold value or less. .
  • the shortage capacity is a cache capacity necessary for the cache capacity of the cache write waiting rate to be equal to or less than the caution value.
  • FIG. 9 shows an example of the attention value excess target table 306.
  • the attention value excess target table 306 stores information on the attention value excess cache division region (target part cache), the MAX value of the cache write wait rate, and the insufficient capacity in association with each other.
  • This insufficient capacity is a cache capacity that needs to be added to the cache partition area that exceeds the caution value, and the cache capacity that is necessary for the MAX value of the cache write wait rate to be at least a predetermined cache capacity that is equal to or less than the cache write wait rate. It is.
  • the shortage capacity is a cache capacity necessary for the cache capacity of the cache write waiting rate to be equal to or less than the caution value.
  • FIG. 10 shows an example of the rearrangement information table 307.
  • the relocation information table 307 information on the difference between the MAX value of the cache write wait rate and the caution value, the deleteable cache capacity, and the delete capacity indicating the cache capacity to be deleted corresponds to each cache partition area. Attached and stored.
  • the relocation review unit 203 when the relocation review unit 203 is considering relocation for a cache partition area that exceeds the threshold, information on the case where the threshold is exceeded is stored.
  • the unit 203 is considering relocation of the cache partition area exceeding the caution value, information when the caution value is exceeded is stored.
  • the cache capacity that can be deleted is the cache capacity that can be deleted from the cache partition area where the cache capacity is sufficient.
  • a cache partition area with sufficient cache capacity is a cache partition area where the cache write wait rate is less than the threshold value (less than the caution value in this example) when the threshold value is exceeded.
  • the cache write waiting rate is at least a cache capacity that is less than the cache write wait rate (in this example, less than the caution value).
  • FIG. 11 shows an example of the storage cache configuration information table 308.
  • the storage cache configuration information table 308 stores information on the cache capacity (mounting capacity) of the entire cache memory 34.
  • the storage cache configuration information table 308 stores a cache capacity (allocated capacity) allocated for each cache partition area.
  • FIG. 12 shows an example of the processing contents (overall flow) related to leveling the load of the cache partition area.
  • the monitoring management unit 202 starts acquiring monitoring information (cache write waiting rate for each cache division area) at every monitoring interval (step S1), and moves the process to step S2. Note that the monitoring management unit 202 continues to acquire the monitoring information even after moving the process to step S2.
  • step S2 the monitoring management unit 202 acquires a reference value from each monitoring item. More specifically, when the regular check interval elapses, the monitoring management unit 202 acquires the time during which the cache write wait rate for the regular check interval is monitored and the cache write wait rate from the cache monitoring data table 304. Next, the monitoring management unit 202 specifies the MAX value of the cache write waiting rate for each cache division area, and counts the number of times that the threshold value has been exceeded and the number of times that the attention value has been exceeded. Then, the MAX value of the specified cache write wait rate and the counted number of times of exceeding the threshold and the number of times of exceeding the attention value are stored in the reference value table 303.
  • the threshold value management unit 201 compares the MAX value of each cache write waiting rate in the reference value table 303 with the threshold value in the threshold value management table 301, and exceeds the threshold value exceeded table 305 and the attention value value.
  • the target table 306 is updated (step S3). More specifically, the threshold value management unit 201 compares the cache write waiting rate MAX value of each cache divided area with the threshold value, and the cache write waiting rate MAX value of the cache divided area exceeding the threshold value. Is added to the threshold exceeded target table 305. The insufficient capacity is added (registered) in step S5 described later.
  • the threshold management unit 201 compares the MAX value of the cache write waiting rate of each cache partition area with the threshold value and the caution value, and the cache write of the cache partition area that exceeds the caution value but falls below the threshold value.
  • the MAX value of the waiting rate is added to the attention value excess target table 306.
  • the relocation examining unit 203 determines whether or not there is a cache partition area in which the MAX value of the cache write wait rate exceeds the threshold value (step S4). More specifically, the reallocation review unit 203 refers to the threshold value excess target table 305 and determines whether information (record) is stored.
  • the difference between the MAX value of the cache write wait rate and the threshold value is calculated, and the processing (steps S5 to S9) is sequentially performed from the cache division area where the difference is large.
  • a cache divided area having a large number of times of exceeding the threshold is preferentially processed.
  • the rearrangement of the cache divided area having a large difference between the MAX value of the cache write waiting rate and the threshold value is preferentially performed, so that a response to the corresponding server 2 is received due to the high load of the cache divided area. A situation such as waiting can be avoided.
  • the administrator may change the priority order of the difference between the MAX value of the cache write wait rate and the threshold value and the number of times the threshold value is exceeded. For example, when the priority of the number of times the threshold is exceeded is made higher than the difference between the MAX value of the cache write wait rate and the threshold, it is possible to avoid a situation where the response to the server 2 is frequently slowed down.
  • step S5 If the relocation examining unit 203 determines that there is a cache partition area in which the MAX value of the cache write wait rate exceeds the threshold, the process moves to step S5, and the MAX value of the cache write wait rate exceeds the threshold. If it is determined that there is no cache division area, the process proceeds to step S2.
  • step S5 the reallocation review unit 203 calculates an insufficient cache capacity (insufficient capacity). More specifically, the relocation review unit 203 calculates a cache capacity necessary for the MAX value of the cache write wait rate to be equal to or less than the caution value.
  • the reallocation review unit 203 refers to the storage cache configuration information table 308, acquires the current allocated capacity, and subtracts the allocated capacity from the necessary cache capacity, thereby calculating the insufficient cache capacity. Then, the reallocation review unit 203 stores the calculated insufficient cache capacity in the threshold value excess target table 305.
  • the MAX value of the cache write waiting rate of the first cache partition area is 55%
  • the allocated capacity of the first cache partition area is 100 GB as shown in FIG.
  • a case where the cache capacity necessary to be equal to or less than the indicated attention value (30%) is calculated will be described as an example.
  • the relocation examining unit 203 determines whether there is room in other cache division areas (step S6). More specifically, the reallocation review unit 203 compensates for the insufficient cache capacity with one or a plurality of other cache division areas that are cache division areas whose MAX value of the cache write wait rate is not more than a caution value. It is determined whether or not The details of this determination will be described with reference to FIG.
  • the relocation review unit 203 moves the process to step S8 if it determines that there is room in another cache partition area, and moves the process to step S7 if it determines that there is no room in another cache partition area.
  • step S7 the reallocation review unit 203 notifies an administrator (such as a management terminal (not shown) capable of communicating with the storage device 3) that the cache memory 34 has no room, and ends the process.
  • an administrator such as a management terminal (not shown) capable of communicating with the storage device 3) that the cache memory 34 has no room, and ends the process.
  • step S8 the configuration changing unit 204 rearranges the cache partition areas. More specifically, the configuration changing unit 204 deletes the cache capacity from the spare cache divided area and adds the cache capacity to the insufficient cache divided area. Details of this processing will be described with reference to FIG.
  • the reallocation review unit 203 determines whether there is a cache partition area in which the MAX value of the cache write wait rate exceeds a threshold value, that is, whether there is a cache partition area to be processed remaining (step) S9). If the relocation examining unit 203 determines that there is a cache partition area in which the MAX value of the cache write wait rate exceeds the threshold, the process moves to step S5, and the MAX value of the cache write wait rate exceeds the threshold. If it is determined that there is no cache division area, the process proceeds to step S10.
  • step S10 the relocation examining unit 203 determines whether or not there is a cache partition area in which the MAX value of the cache write wait rate exceeds the caution value but falls below the threshold value. More specifically, the rearrangement review unit 203 refers to the caution value excess target table 306 and determines whether information (record) is stored. At this time, when a plurality of records are stored, the difference between the MAX value of the cache write waiting rate and the caution value is calculated, and the cache division area with the largest difference is processed (steps S11 to S13). To do. In addition, when there are a plurality of cache divided areas having the same difference, the cache divided areas having a large number of times the warning value is exceeded are preferentially processed.
  • step S11 the MAX value of the cache write wait rate If it is determined that there is no cache partition area that exceeds the caution value but is below the threshold value, the process proceeds to step S2.
  • the load of the cache division area exceeding the caution value is further leveled compared to the case of relocation of only the cache division area where the difference between the MAX value of the cache write waiting rate and the caution value is the maximum.
  • the relocation review unit 203 determines whether there is room in another cache partition area (step S11). More specifically, the rearrangement review unit 203 is another cache partition area that is different from the cache partition area to be processed, and is a cache partition area in which the MAX value of the cache write wait rate is a caution value or less. It is determined whether or not a predetermined cache capacity can be supplemented by a plurality of cache division areas.
  • the relocation review unit 203 moves the process to step S13 if it determines that there is room in another cache partition area, and moves the process to step S12 if it determines that there is no room in another cache partition area. Note that the details of this determination are the same as the processing shown in FIG. 13, and the threshold value exceeded target table 305 is mainly replaced with the caution value exceeded target table 306, and thus description thereof is omitted.
  • step S12 the reallocation review unit 203 notifies an administrator (such as a management terminal (not shown) capable of communicating with the storage apparatus 3) that the cache memory 34 has no room, and the process proceeds to step S14. .
  • an administrator such as a management terminal (not shown) capable of communicating with the storage apparatus 3
  • step S13 the configuration changing unit 204 rearranges the cache partition areas. More specifically, the configuration changing unit 204 deletes the cache capacity from the spare cache divided area and adds the cache capacity to the processing target cache divided area. Details of this processing will be described with reference to FIG.
  • step S10 to step S13 it is confirmed whether there is a cache partition area (prevention target cache partition area) in which the cache write wait rate exceeds the caution value but is less than the threshold value. If there is a cache partition area to be prevented, the additional capacity (insufficient capacity) required in the cache partition area to be prevented is calculated, and it is determined whether there is room in other cache partition areas. Relocation of the cache partition area is performed. According to such processing, the load on the cache partition area is further leveled, and consequently, leveled proactively.
  • step S14 the rearrangement review unit 203 creates a report, outputs the report to an administrator (such as a management terminal (not shown) that can communicate with the storage device 3), and moves the processing to step S2.
  • the relocation examination unit 203 performs the processing result, information indicating the cache partition area to be noted (for example, information on the number of times the cache partition area exceeds the caution value), and information related to the cache partition area having a sufficient cache capacity ( Include cache capacity that can be deleted, information on deleted capacity, etc.) in the report.
  • FIG. 13 shows an example of the processing contents related to the determination as to whether or not there is room in other cache division areas.
  • the rearrangement review unit 203 calculates a deletable cache capacity (deletable capacity) for all cache divided areas in the reference value table 303 (step S21). More specifically, the reallocation review unit 203 calculates the difference between the MAX value of the cache write wait rate and the caution value for all the cache divided areas in the reference value table 303, and calculates the deleteable cache capacity.
  • the relocation information table 307 is updated.
  • the attention value is 30% as shown in FIG. 4
  • the MAX value of the cache write wait rate in the second cache partition area is 15% as shown in FIG. 6, and the second value as shown in FIG.
  • a method for calculating the removable cache capacity will be described by taking as an example a case where the allocated capacity of the cache divided area is 400 GB.
  • the relocation examining unit 203 determines whether or not the insufficient capacity is equal to or less than the total amount of removable capacity (step S22). If the relocation examining unit 203 determines that the shortage capacity is less than or equal to the total amount of deleteable capacity, it moves the process to step S23, and if it determines that the shortage capacity is not less than or equal to the total amount of deleteable capacity, It is determined that there is no room, and the process is moved to the entire flow.
  • the relocation examining unit 203 determines whether or not there is one cache partition area having a deleteable capacity (step S23). If the relocation examination unit 203 determines that there is one cache partition area having a deleteable capacity, it moves the process to step S24, and if it determines that there is not one cache partition area having a deleteable capacity, the step The processing is moved to S25.
  • step S24 the rearrangement review unit 203 updates the deleteable cache capacity and the deletion capacity of the rearrangement information table 307, determines that there is room in other cache division areas, and shifts the processing to the entire flow.
  • step S25 the reallocation review unit 203 refers to the reallocation information table 307 and selects a cache division area (cache division area A) having the largest difference between the MAX value of the cache write wait rate and the caution value as a processing target. To do.
  • the relocation examining unit 203 calculates a cache capacity for increasing the cache write wait rate by a predetermined amount (for example, 1%) and sets it as a deletion capacity (step S26).
  • the relocation examining unit 203 updates the difference between the MAX value of the cache write wait rate and the caution value and the deletion capacity in the relocation information table 307 (step S27).
  • the relocation examining unit 203 determines whether or not the shortage capacity is equal to or less than the total amount of deletion capacity (step S28). When the rearrangement examining unit 203 determines that the shortage capacity is equal to or less than the total amount of deletion capacity, the relocation examining unit 203 proceeds to step S29, and when it is determined that the shortage capacity is not equal to or less than the total amount of deletion capacity, the processing is shifted to step S30.
  • step S29 the reallocation review unit 203 updates the cache capacity that can be deleted from the reallocation information table 307, determines that there is room in another cache division area, and moves the processing to the entire flow.
  • step S30 the reallocation review unit 203 differs from the cache partition area having the next largest cache write wait rate after the cache partition area A even if the cache write wait ratio of the cache partition area A is increased by a predetermined amount (for example, 1%). It is determined whether or not there is. If the rearrangement examining unit 203 determines that there is a difference, it moves the process to step S26, and if it determines that there is no difference, it moves the process to step S31.
  • a predetermined amount for example, 1%
  • step S31 the reallocation review unit 203 sets the cache partition area A and the cache partition area having the next largest difference in the cache write wait rate after the cache partition area A as the cache partition area A, and moves the process to step S25. .
  • step S25 to step S30 will be described by taking the case of the above as an example.
  • step S ⁇ b> 30 the reallocation review unit 203 finds that the difference between the eighth cache divided area having the largest cache write wait rate MAX value next to the seventh cache divided area and the seventh cache divided area having the cache write wait rate of 11% is Since there is no cache write waiting ratio, the eighth cache partition area having a MAX value of 11% is set as a processing target (cache partition area A) (step S31).
  • the reallocation review unit 203 moves the process to step S30 because the shortage capacity 20 GB cannot be satisfied with the total amount of deletion capacity 9 GB.
  • the reallocation review unit 203 uses the seventh cache divided area and the eighth cache divided area, the ninth cache divided area having the largest MAX value of the cache write wait rate, and the seventh cache with the cache write wait rate set to 12%. Since there is no difference between the divided area and the eighth cache divided area, the seventh cache divided area with a cache write wait rate of 12%, the eighth cache divided area with a cache write wait rate of 12%, and the cache write wait
  • the ninth cache partition area whose rate MAX value is 12% is set as a processing target (cache partition area A) (step S31).
  • the process is repeated in this way, and the relocation examination unit 203 moves the process to step S29 when a deletion capacity sufficient to satisfy the shortage capacity 20 GB can be secured.
  • FIG. 14 shows an example of the processing contents related to the rearrangement of the cache division area.
  • the configuration changing unit 204 performs capacity change (capacity deletion) for a cache division area having a deletion capacity (step S41). More specifically, the configuration changing unit 204 refers to the relocation information table 307, identifies one cache partition area having a deleted capacity, and extracts the deleted capacity. Next, the configuration change unit 204 refers to the storage cache configuration information table 308, identifies the allocated capacity of the cache partition area, and sets a value obtained by subtracting the deleted capacity extracted above from the identified allocated capacity as a new allocated capacity. The storage cache configuration information table 308 is updated. Next, the configuration change unit 204 instructs the cache memory management unit 205 to rearrange the cache divided areas.
  • capacity change capacity deletion
  • the cache memory management unit 205 allocates the address of the storage area of the cache memory 34 corresponding to the cache divided area so as to reduce the cache divided area (delete the cache capacity) according to the allocated capacity of the storage cache configuration information table 308. To change. At this time, if there is data in the cache divided area, the cache memory management unit 205 changes the allocation of the address of the storage area after saving the data.
  • the configuration changing unit 204 determines whether or not there is an unprocessed cache divided area having a deleted capacity (step S42). If the configuration changing unit 204 determines that there is an unprocessed cache divided area having a deleted capacity, the process is shifted to step S41. If it is determined that there is no unprocessed cache divided area having a deleted capacity, the process is changed to step S43. Move processing.
  • the configuration changing unit 204 changes the capacity of the cache division area exceeding the threshold (adds capacity) (step S43), and shifts the processing to the entire flow. More specifically, the configuration changing unit 204 refers to the storage cache configuration information table 308, identifies the allocated capacity of the cache partition area that exceeds the threshold, and deletes the calculated allocated capacity from the relocation information table 307. The storage cache configuration information table 308 is updated with the sum of the capacities as a new allocated capacity. Next, the configuration change unit 204 instructs the cache memory management unit 205 to rearrange the cache divided areas.
  • the cache memory management unit 205 stores the cache memory 34 corresponding to the threshold-exceeded cache partition area so as to expand the cache partition area exceeding the threshold (add cache capacity) according to the allocated capacity of the storage cache configuration information table 308. Change the storage address assignment.
  • the configuration changing unit 204 updates the rearrangement information table 307 with the deletion capacity of the cache partition area set to “0” at an appropriate timing.
  • the allocation of the cache capacity of the cache partition area is revised so that the load of the cache partition area is leveled based on the operation status after the operation of the computer system 1. Therefore, the cache memory 34 can be used effectively.
  • the present invention is not limited to this, and when the relocation examining unit 203 determines that there is no cache division region where the MAX value of the cache write wait rate exceeds the threshold value, the processing may be shifted to step S10. . According to this configuration, the load on the cache memory 34 can be leveled proactively even when there is no cache division area in which the MAX value of the cache write wait rate exceeds the threshold value.
  • the difference between the MAX value of the cache write wait rate and the threshold value is calculated, and the processing is sequentially performed from the cache divided area having the large difference.
  • the cache divided area having a large number of times the threshold is exceeded is preferentially processed, but the present invention is not limited to this, and cache writing You may make it the object of a process sequentially from the thing with the short monitoring time (time) of the MAX value of a waiting rate, and may combine these suitably.
  • the reallocation review unit 203 determines whether or not there is one cache partition area having a erasable capacity has been described.
  • the arrangement review unit 203 may determine whether there is a cache partition area having a removable capacity larger than the shortage capacity.
  • the reallocation review unit 203 has insufficient cache capacity that needs to be added because the cache write wait information of the specified cache partition area is below a pre-defined caution value that is lower than the threshold value. It is calculated as the cache capacity.
  • the reallocation review unit 203 determines, based on the cache capacity that the cache write wait information of each cache partition area becomes a caution value among the cache partition areas other than the cache partition area identified as exceeding the threshold value.
  • the reallocation review unit 203 compares each of the insufficient cache capacity and each of the deleteable cache capacity, and determines whether or not the cache capacity can be rearranged. According to such a configuration, the insufficient cache capacity can be compensated with one cache division area.
  • the present invention is not limited to this, and the average cache for a certain period is described. Information such as a write wait rate and a ratio that can be calculated from a cache write wait rate for a certain period of time may be acquired.
  • the present invention is not limited to this, and may be a cache capacity necessary for an average value of a predetermined period such as a periodic check interval of the cache write wait rate to be a predetermined cache capacity that is at least a threshold value or less.
  • the cache capacity may be a cache capacity necessary for the average value of the plurality of cache write waiting ratios to be a predetermined cache capacity that is at least a threshold value or less.
  • the predetermined cache capacity may be a cache capacity in which the MAX value of the cache write wait rate is not more than a caution value, or may be a cache capacity in which the average value of the cache write wait rate for a predetermined period is not more than a threshold value.
  • other cache capacity may be used.
  • the predetermined cache capacity may be changed by the administrator.
  • the present invention is not limited to this, and may be a cache capacity necessary for an average value of a predetermined period such as a periodic check interval of the cache write wait rate to be at least a predetermined cache capacity equal to or less than the cache write wait rate.
  • the cache capacity required for the average value of the plurality of cache write waiting ratios to be a predetermined cache capacity at least equal to or less than the cache write waiting ratio may be used.
  • the predetermined cache capacity may be a cache capacity in which the MAX value of the cache write wait rate is equal to or less than the caution value, or may be a cache capacity in which the average value of the cache write wait rate for a predetermined period is equal to or less than the caution value. Other cache capacities may be used.
  • the predetermined cache capacity may be changed by the administrator.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Memory System Of A Hierarchy Structure (AREA)

Abstract

L'invention concerne un dispositif de mémorisation comprenant une unité d'examen de réattribution qui : identifie, parmi une pluralité de zones de mémoire cache divisées, une zone de mémoire cache divisée qui dépasse un seuil prédéterminé, en fonction d'informations d'attente d'écriture de mémoire cache surveillées concernant chacune des zones de mémoire cache divisées ; calcule la quantité manquante à la zone de mémoire cache divisée identifiée par rapport à la capacité de mémoire cache exigée ; calcule la quantité de capacité de mémoire cache pouvant être désattribuée dans chaque zone de mémoire cache divisée autre que la zone de mémoire cache divisée identifiée ; et compare la quantité calculée manquant à la zone de mémoire cache divisée identifiée par rapport à la capacité de mémoire cache exigée avec la quantité calculée de capacité de mémoire cache pouvant être désattribuée dans chaque zone de mémoire cache divisée, et détermine ainsi s'il est possible ou non de réattribuer une capacité de mémoire cache à des zones de mémoire cache divisées. Ledit dispositif de mémorisation comprend en outre une unité de changement de configuration qui réattribue une capacité de mémoire cache à la pluralité de zones de mémoire cache divisées en fonction de résultats de détermination provenant de l'unité d'examen de réattribution.
PCT/JP2017/015040 2017-04-12 2017-04-12 Dispositif de mémorisation et procédé de gestion de mémoire cache WO2018189847A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/JP2017/015040 WO2018189847A1 (fr) 2017-04-12 2017-04-12 Dispositif de mémorisation et procédé de gestion de mémoire cache

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/JP2017/015040 WO2018189847A1 (fr) 2017-04-12 2017-04-12 Dispositif de mémorisation et procédé de gestion de mémoire cache

Publications (1)

Publication Number Publication Date
WO2018189847A1 true WO2018189847A1 (fr) 2018-10-18

Family

ID=63792950

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2017/015040 WO2018189847A1 (fr) 2017-04-12 2017-04-12 Dispositif de mémorisation et procédé de gestion de mémoire cache

Country Status (1)

Country Link
WO (1) WO2018189847A1 (fr)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115952110A (zh) * 2023-03-09 2023-04-11 浪潮电子信息产业股份有限公司 一种数据缓存方法、装置、设备和计算机可读存储介质
US11687255B2 (en) 2021-01-08 2023-06-27 Fujitsu Limited Information processing system and information processing method

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013510360A (ja) * 2010-03-19 2013-03-21 株式会社日立製作所 ストレージシステム及びストレージシステムにおけるキャッシュの構成変更方法
WO2014178104A1 (fr) * 2013-04-30 2014-11-06 株式会社日立製作所 Système et procédé informatiques d'aide à l'analyse d'une reproduction asynchrone à distance
US20140359226A1 (en) * 2013-05-30 2014-12-04 Hewlett-Packard Development Company, L.P. Allocation of cache to storage volumes
JP2015060376A (ja) * 2013-09-18 2015-03-30 富士通株式会社 キャッシュメモリ制御プログラム，キャッシュメモリを内蔵するプロセッサ及びキャッシュメモリ制御方法
WO2015045046A1 (fr) * 2013-09-26 2015-04-02 株式会社日立製作所 Système informatique et procédé de réglage d'allocation de mémoire pour un système informatique

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2013510360A (ja) * 2010-03-19 2013-03-21 株式会社日立製作所 ストレージシステム及びストレージシステムにおけるキャッシュの構成変更方法
WO2014178104A1 (fr) * 2013-04-30 2014-11-06 株式会社日立製作所 Système et procédé informatiques d'aide à l'analyse d'une reproduction asynchrone à distance
US20140359226A1 (en) * 2013-05-30 2014-12-04 Hewlett-Packard Development Company, L.P. Allocation of cache to storage volumes
JP2015060376A (ja) * 2013-09-18 2015-03-30 富士通株式会社 キャッシュメモリ制御プログラム，キャッシュメモリを内蔵するプロセッサ及びキャッシュメモリ制御方法
WO2015045046A1 (fr) * 2013-09-26 2015-04-02 株式会社日立製作所 Système informatique et procédé de réglage d'allocation de mémoire pour un système informatique

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11687255B2 (en) 2021-01-08 2023-06-27 Fujitsu Limited Information processing system and information processing method
CN115952110A (zh) * 2023-03-09 2023-04-11 浪潮电子信息产业股份有限公司 一种数据缓存方法、装置、设备和计算机可读存储介质

Similar Documents

Publication Publication Date Title
US11372544B2 (en) Write type based crediting for block level write throttling to control impact to read input/output operations
US11150829B2 (en) Storage system and data control method
US7415573B2 (en) Storage system and storage control method
US9823875B2 (en) Transparent hybrid data storage
JP5439581B2 (ja) ストレージシステム、ストレージ装置、ストレージシステムの記憶領域の最適化方法
US20080229048A1 (en) Method and apparatus for chunk allocation in a thin provisioning storage system
US7975123B2 (en) Computer system, management computer and storage system, and storage area allocation amount controlling method
US8924681B1 (en) Systems, methods, and computer readable media for an adaptative block allocation mechanism
JP6340439B2 (ja) ストレージシステム
JP5228988B2 (ja) 割当制御プログラム及び割当制御装置
CN109074226A (zh) 一种存储系统中重复数据删除方法、存储系统及控制器
US10372372B2 (en) Storage system
US9513837B2 (en) Performance assist storage volumes
JP2012533781A (ja) 計算機システム及びその負荷均等化制御方法
US7337283B2 (en) Method and system for managing storage reservation
US9734048B2 (en) Storage management device, performance adjustment method, and computer-readable recording medium
WO2018189847A1 (fr) Dispositif de mémorisation et procédé de gestion de mémoire cache
US20070079103A1 (en) Method for resource management in a logically partitioned storage system
US8984235B2 (en) Storage apparatus and control method for storage apparatus
US20190332261A1 (en) Storage system, method of controlling storage system, and management node
US9547443B2 (en) Method and apparatus to pin page based on server state
CN117389485B (zh) 存储性能优化方法、装置、存储系统、电子设备和介质
WO2010119478A1 (fr) Système de stockage et son procédé de commande
WO2018116392A1 (fr) Système de traitement d'informations et procédé de traitement d'informations

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17905168

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17905168

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP