WO2018061158A1

WO2018061158A1 - Computer system and computer system control method

Info

Publication number: WO2018061158A1
Application number: PCT/JP2016/078912
Authority: WO
Inventors: 寛之林; 貴司戸高
Original assignee: 株式会社日立製作所
Priority date: 2016-09-29
Filing date: 2016-09-29
Publication date: 2018-04-05

Abstract

A first host processor executes a first in-memory database to write data in the first in-memory database to a first memory and record an access log of the first memory. A first management processor generates access information indicating data accessed within a specific period by the first in-memory database in the first memory on the basis of the access log. A first storage processor controls a first storage tier that uses a first upper storing device and a first lower storing device, and disposes data written to the first memory within the specific period in the first upper storing device on the basis of the access information.

Description

Computer system and computer system control method

The present invention relates to a computer system.

Recently, in order to cope with new business such as big data / IoT (Internet of Things) and rapidly changing business environment, the demand for real-time processing is increasing, and interest in in-memory databases is increasing.

Patent Document 1 describes tiering using a memory and a disk.

US Pat. No. 9,189,387

The in-memory database periodically backs up the data in the host server memory to the storage device. In-memory databases have the problem of performance degradation. For example, when data is written to the HDD during backup and the write data overflows from the storage cache, the response time is deteriorated.

In order to solve the above problems, a computer system according to an aspect of the present invention includes a first memory, a first host processor connected to the first memory, a first upper storage device, and the first upper storage. A first lower storage device having performance lower than that of the device, the first host processor, the first upper storage device, a first storage processor connected to the first lower storage device, and the first host processor And a first management processor connected to the first storage processor. The first host processor executes the first in-memory database, writes the data in the first in-memory database to the first memory, and records the access log of the first memory. The first management processor generates access information indicating data accessed in the first memory by the first in-memory database within a specific period based on the access log. The first storage processor controls a first storage tier that uses the first upper storage device and the first lower storage device, and is written to the first memory within the specific period based on the access information. Is stored in the first upper storage device.

∙ Performance degradation in the in-memory database can be prevented.

An outline of the operation of the computer system of this embodiment will be shown. The configuration of the computer system is shown. An access log table 710 is shown. A memory mapping table 720 is shown. A database information table 730 is shown. An SLA information table 740 is shown. A storage information table 750 is shown. A statistical information table 760 is shown. A statistical information generation process is shown. Indicates the information collection process. The main site data arrangement process is shown.

Hereinafter, embodiments of the present invention will be described with reference to the drawings.

In the following description, information may be described using the expression “xxx table”, but the information may be expressed in any data structure. That is, “xxx table” can be referred to as “xxx information” to indicate that the information does not depend on the data structure. In the following description, the configuration of each table is an example, and one table may be divided into two or more tables, or all or part of the two or more tables may be a single table. Good.

In the following description, an ID is used as element identification information, but other types of identification information may be used instead of or in addition thereto.

In the following description, when a description is made without distinguishing the same type of element, a reference number or a common number in the reference number is used, and when a description is made by distinguishing the same type of element, the reference number of the element is used. Alternatively, an ID assigned to the element may be used instead of the reference code.

In the following description, an I / O (Input / Output) request is a write request or a read request, and may be referred to as an access request.

In the following description, the process may be described using “program” as a subject. However, the program is executed by a processor (for example, a CPU (Central Processing Unit)), so that a predetermined processing is appropriately performed. Since processing is performed using a storage resource (for example, a memory) and / or an interface device (for example, a communication port), the subject of processing may be a processor. The process described with the program as the subject may be a process or system performed by a processor or an apparatus having the processor. The processor may include a hardware circuit that performs a part or all of the processing. The program may be installed in a computer-like device from a program source. The program source may be, for example, a storage medium that can be read by a program distribution server or a computer. When the program source is a program distribution server, the program distribution server may include a processor (for example, a CPU) and a storage resource, and the storage resource may further store a distribution program and a program to be distributed. Then, the processor of the program distribution server executes the distribution program, so that the processor of the program distribution server may distribute the distribution target program to other computers. In the following description, two or more programs may be realized as one program, or one program may be realized as two or more programs.

In the following description, the management system may include one or more computers. Specifically, for example, when the management computer displays information (specifically, for example, the management computer displays information on its own display device, or the management computer displays display information in a remote display computer) Management computer is the management system. For example, when a function equivalent to that of the management computer is realized by a plurality of computers, the plurality of computers (may include a display computer when the display computer performs display) is the management system. . The management computer (eg, management system) may include an interface device connected to the I / O system including the display system, a storage resource (eg, memory), and a processor connected to the interface device and the storage resource. The display system may be a display device included in the management computer or a display computer connected to the management computer. The I / O system may be an I / O device (for example, a keyboard and a pointing device or a touch panel) included in the management computer, a display computer connected to the management computer, or another computer. “Displaying display information” by the management computer means displaying the display information on the display system, which may be displaying the display information on a display device included in the management computer. The management computer may transmit display information to the display computer (in the latter case, the display information is displayed by the display computer). The management computer inputting / outputting information may be inputting / outputting information to / from an I / O device of the management computer, or a remote computer connected to the management computer (for example, a display) Information may be input / output to / from the computer. The information output may be a display of information.

FIG. 1 shows an outline of the operation of the computer system of this embodiment.

The computer system of this embodiment includes a primary site 10 and a secondary site 20. The primary site 10 includes a host server 100, a storage device 200, and a management server 400. The secondary site 20 includes a host server 100, a storage device 200, and a management server 400. Each storage device 200 includes a high-speed storage device 221 and a low-speed storage device 222.

The first problem related to performance degradation is that when data in the memory 110 of the host server 100 is backed up to the storage device 200, data write to the HDD occurs, and if the write data overflows from the storage cache, the response time deteriorates. It is to be. The computer system of this embodiment can solve this problem by the following operation.

In the primary site 10, the host server 100 executes the DB 500, which is an instance of the in-memory database. The memory 110 of the host server 100 stores a log 510 and data 520 of the DB 500. The host server 100 writes the log 510 to the storage apparatus 200 each time the DB 500 commits. Further, the host server 100 transmits the data 520 written by the DB 500 to the storage apparatus 200 every time a preset data backup time interval elapses.

In the primary site 10, the management server 400 generates the statistical information 540 based on the information collected from the host server 100 and the storage device 200 every time a preset statistical information generation time interval elapses. Transmit to the storage device 200. It is desirable that the statistical information generation time interval is shorter than the data backup time interval.

At the primary site 10, the host server 100 performs log backup by writing the log of the DB 500 in the memory 110 to the storage device 200. Further, the host server 100 performs data backup by writing the data of the DB 500 in the memory 110 to the storage device 200. The storage apparatus 200 classifies data from the host server 100 into hot data 620 and cold data 630, places the hot data 620 in the high speed storage device 221, and places the cold data 630 in the low speed storage device 222. Further, the storage apparatus 200 places the log 610 in the high-speed storage device 221.

At the primary site 10, the storage apparatus 200 updates the log 610 with the log 510 received from the host server 100. Based on the statistical information 640 received from the host server 100, the storage apparatus 200 determines the data received from the host server 100 as the Hot data 620, determines the other data as Cold data 630, and arranges the data To do.

When the host server 100 executes a plurality of instances of the in-memory database, the host server 100 performs data backup at a data backup time interval for each instance. The host server 100 sets a scheduled data backup time for each instance at a data backup time interval, and performs data backup of the instance at the scheduled data backup time of the instance.

Through the above operation, the host server 100 of the primary site 10 periodically writes the updated data to the high-speed storage device 221. Thereby, deterioration of response time can be prevented.

A second problem related to performance degradation is that when data in the storage device 200 at the primary site 10 is copied to the storage device 200 at the secondary site 20 using the storage copy function as a disaster countermeasure, the secondary site 20, it is necessary to load the data in the storage apparatus 200 into the memory 110 of the host server 100, which increases RTO (recovery time objective). The computer system of this embodiment can solve this problem by the following operation.

The storage device 200 at the primary site 10 transmits the statistical information 640 to the storage device 200 at the secondary site 20. Furthermore, the storage device 200 at the primary site 10 performs replication by transmitting the update amount of the log 610 and the update amount of the data to the storage device 200 of the secondary site 20.

In the secondary site 20, the storage apparatus 200 determines that the data accessed (read or written) by the host server 100 of the primary site 10 is the Hot data 620 based on the statistical information 640, and the other data is Cold data 630. Is determined and data is arranged.

When failover from the primary site 10 to the secondary site 20 is performed due to a failure of the primary site 10, the host server 100 stores the log 610 and the hot data 620 stored in the high-speed storage device 221 in the secondary site 20. By loading to 110, the DB 500 is restored.

By the above operation, the host server 100 of the secondary site 20 loads the log 610 and the hot data 620 from the high-speed storage device 221 to the memory 110 at the time of recovery. Thereby, RTO can be shortened.

Fig. 2 shows the configuration of the computer system.

The primary site 10 and the secondary site 20 are connected to each other via a communication network. The primary site 10 includes a host server 100 (host server # 1-1), a storage device 200 (storage device # 1), and a management server 400 (management server # 1). Similar to the primary site 10, the secondary site 20 includes a host server 100 (host server # 2-1), a storage device 200 (storage device # 2), and a management server 400 (management server # 2). The host server 100, the storage device 200, and the management server 400 are connected to each other via a communication network. The primary site 10 may include a plurality of host servers 100 (host servers # 1-1, # 1-2,..., # 1-N). The secondary site 20 may include a plurality of host servers 100 (host servers # 2-1, # 2-2,..., # 2-N).

The host server 100 includes a memory 110 and a processor 120. The memory 110 stores programs and data. The memory 110 stores a DBMS (database management system) 111 that is a program for managing an in-memory database, and an OS 112. The processor 120 performs processing according to a program stored in the memory 110.

The storage device 200 includes a controller 210 and a recording unit 220.

The controller 210 includes a memory 310 and a processor 320. The memory 310 stores programs and data. The memory 310 stores a hierarchy control unit 311 for controlling the hierarchy of the recording unit 220. The processor 320 performs processing according to a program stored in the memory 310.

The recording unit 220 includes a high-speed storage device 221 and a low-speed storage device 222. The access performance of the high speed storage device 221 is higher than the access performance of the low speed storage device 222. The high-speed storage device 221 in the present embodiment is a flash drive (SSD: solid state drive). The low-speed storage device 222 in this embodiment is an HDD (hard disk drive). The tier control unit 311 assigns the high speed storage device 221 and the low speed storage device 222 to two tiers, respectively.

The management server 400 includes a memory 410 and a processor 420. The memory 410 stores programs and data. The memory 410 includes an access log collection unit 411, a memory mapping collection unit 412, a database information collection unit 413, an SLA (service level agreement) information collection unit 414, a storage information collection unit 415, and a data statistics analysis unit 416. Is stored. The processor 420 performs processing according to a program stored in the memory 410.

Note that each of the primary site 10 and the secondary site 20 may not include the management server 400. In this case, the access log collection unit 411, the memory mapping collection unit 412, the database information collection unit 413, the SLA information collection unit 414, the storage information collection unit 415, and the data statistics analysis unit 416 are the host server 100 and It may be stored in any storage device 200 and executed. Further, the host server 100 and the storage apparatus 200 may be one computer.

Hereinafter, information generated by the management server 400 will be described.

The host server 100 sets a monitoring period (specific period) after the previous data backup time, and records an access log indicating memory access to the memory 110 within the monitoring period. The access log collection unit 411 acquires an access log from the host server 100 and records the access log in the access log table 710.

FIG. 3 shows the access log table 710.

The access log table 710 has an entry for each memory access. One memory access entry includes a time 711 of the memory access, an R / W 712 indicating whether the type of the memory access is Read or Write, and a physical in the memory 110 that is a target of the memory access. Address 713.

The host server 100 assigns a physical address in the memory 110 to a virtual address set for memory access, and records the assignment as memory mapping. The memory mapping collection unit 412 acquires memory mapping from the host server 100 and records the memory mapping in the memory mapping table 720.

FIG. 4 shows the memory mapping table 720.

The memory mapping table 720 has an entry for each physical address. The entry of one physical address includes a physical address 721 indicating the physical address and a virtual address 722 associated with the physical address.

The host server 100 records database information indicating load unit data accessed in the monitoring period in the in-memory database. The load unit data is a unit of data loaded from the storage apparatus 200 to the memory 110 at the secondary site at the time of failover. The load unit data may be a unit of data that is backed up from the memory 110 to the storage device 200 at the primary site. The database information collection unit 413 acquires database information from the host server 100 and generates a database information table 730 based on the acquired information. The database information collection unit 413 generates a database information table 730 for each instance of the in-memory database.

FIG. 5 shows the database information table 730.

The database information table 730 has an entry for each load unit data accessed during the monitoring period. One load unit data entry includes a load unit 731 indicating a unit of the load unit data, a virtual address 732 of the load unit data, a disk address 733 which is an address in the storage device 200 of the load unit data, When the load unit data is written, the data backup scheduled time 734, which is the time when the data is written to the storage device 200, is included.

The load unit 731 is, for example, an identifier such as Table or Column. The scheduled data backup time 734 is the next scheduled data backup time and is the end time of the current monitoring period.

The SLA information collection unit 414 acquires the SLA information set by the administrator for each site and each instance of the in-memory database, and generates the SLA information table 740 based on the acquired information.

FIG. 6 shows the SLA information table 740.

The SLA information table 740 has an entry for each instance of the in-memory database. The entry of one instance includes an instance 741 that is an identifier of the instance, an SLA (primary) 742 that indicates the SLA set in the instance at the primary site, and an SLA ( Sub) 743. The SLA is represented by a preset rank, for example. For example, the rank is expressed as Gold, Silver, Blonze in descending order of SLA.

The controller 210 of the storage apparatus 200 records the usage amount of the high-speed storage device 221 for each instance and the free capacity of the high-speed storage device 221 as storage information. The storage information collection unit 415 acquires storage information from the storage apparatus 200 and generates a storage information table 750 based on the acquired information.

FIG. 7 shows the storage information table 750.

The storage information table 750 includes an entry for each instance. The entry of one instance includes an instance 751 that is an identifier of the instance and a high-speed storage device usage amount 752 that is a usage amount of the high-speed storage device 221 by the instance. The storage information table 750 further includes a free capacity entry. An instance 751 of the free capacity entry indicates the free capacity. The high-speed storage device usage amount 752 of the free capacity entry indicates the free capacity of the high-speed storage device 221. The free capacity of the high-speed storage device 221 is a value obtained by subtracting the total of the high-speed storage device usage 752 of all instances from the free capacity of the high-speed storage device 221.

The data statistical analysis unit 416 acquires the access log table 710, the memory mapping table 720, the database information table 730, the SLA information table 740, and the storage information table 750, and based on the acquired information, the statistical information table 760 Is generated. Here, the data statistical analysis unit 416 classifies the acquired information for each instance, each processing type, and each disk address, and calculates an access frequency that is the number of accesses per unit time for the disk address.

FIG. 8 shows the statistical information table 760.

The statistical information table 760 has an entry for each load unit data accessed within the monitoring period. One load unit data entry is set in the instance 761 which is an identifier of the instance of the load unit data, the SLA (primary) 762 which is the SLA set in the instance in the primary site, and the instance in the secondary site. SLA (secondary) 763 that is the SLA, the scheduled data backup time 764 when the load unit data is written, the process 765 indicating the access type (Write or Read) of the load unit data, and the load unit It includes a disk address 766 that is an address of a storage area in the storage apparatus 200 corresponding to the data, and an access frequency 767 of the storage area.

The details of the operation of the computer system will be described below.

The primary site management server 400 executes the statistical information generation process every time the statistical information generation time interval elapses.

FIG. 9 shows statistical information generation processing.

In S110, the management server 400 at the primary site performs information collection processing described later.

In S120, the data statistical analysis unit 416 of the primary site management server 400 creates the statistical information table 760.

In S130, the data statistical analysis unit 416 of the primary site management server 400 transmits the statistical information table 760 to the storage device 200 of the primary site.

In S140, the tier control unit 311 of the primary site storage apparatus 200 performs a primary site data arrangement process described below based on the statistical information table 760.

In S150, the DBMS 111 of the host server 100 at the primary site writes the data in the memory 110 to the storage device 200, and ends this flow.

After S130, in S160, the data statistical analysis unit 416 of the management server 400 at the primary site transmits the statistical information table 760 to the storage device 200 at the secondary site.

In S170, the tier control unit 311 of the storage device 200 at the secondary site performs a secondary site data arrangement process described later based on the statistical information table 760, and ends this flow.

According to the statistical information generation processing described above, the computer system can perform data arrangement of the storage device 200 at the primary site and data arrangement of the storage device 200 at the secondary site based on the statistical information table 760 at the primary site. it can.

In S110 described above, the primary site management server 400 performs information collection processing.

FIG. 10 shows the information collection process.

In S210, the SLA information collection unit 414 acquires the SLA information of each instance and records the SLA information in the SLA information table 740. In S220, the access log collection unit 411 acquires the access log of the memory 110 and records the access log in the access log table 710. In S230, the memory mapping collection unit 412 acquires the memory mapping, and records the memory mapping in the memory mapping table 720. In S240, the database information collection unit 413 acquires the database information and records it in the database information table 730.

According to the above information collection processing, the management server 400 can collect information indicating the state of the host server 100 at the primary site.

In S140 described above, the primary site storage apparatus 200 performs primary site data arrangement processing for each instance in response to reception of the statistical information table 760 from the primary site management server 400.

FIG. 11 shows the primary site data placement process.

In S310, the tier control unit 311 of the primary site storage apparatus 200 acquires the statistical information table 760 from the management server 400.

In S320, the tier control unit 311 of the storage device 200 at the primary site determines whether or not there is a higher placement target instance in the statistical information table 760. For example, the hierarchy control unit 311 determines an instance having an SLA equal to or higher than a specific SLA as a higher placement target instance.

If the result of S320 is No, the hierarchy control unit 311 ends this flow.

If the result of S320 is Yes, in S330, the hierarchy control unit 311 selects one instance among the higher-level placement target instances as the target instance, and the target in the high-speed storage device is selected from the storage information table 750 of the management server 400. The instance usage and free capacity are acquired, and the sum of the usage and free capacity of the target instance is calculated as the usable capacity.

In S340, based on the statistical information table 760, the hierarchy control unit 311 identifies data written to the target instance within the monitoring period as target data, and calculates the target data size.

In S350, the hierarchy control unit 311 determines whether the usable capacity is equal to or larger than the target data size.

If the result of S350 is Yes, in S360, the hierarchy control unit 311 arranges the target data in the high-speed storage device 221 and arranges data other than the target data in the low-speed storage device 222, and ends this flow.

If the result of S350 is No, in S410, the hierarchy control unit 311 determines whether or not the target instance is the next write target. Here, when the scheduled data backup time 764 of the target instance is the earliest among the scheduled data backup times 764 of the plurality of instances in the statistical information table 760, the hierarchy control unit 311 determines that the target instance is the next export target. judge.

When the result of S410 is No, in S440, the hierarchy control unit 311 does not perform data arrangement and ends this flow.

If the result of S410 is Yes, in S420, the hierarchy control unit 311 determines whether the target data size is smaller than the total capacity of the high-speed storage device 221.

If the result of S420 is No, in S440, the hierarchy control unit 311 does not execute data arrangement and ends this flow.

If the result of S420 is Yes, in S430, the hierarchy control unit 311 selects an instance other than the target instance in descending order of the scheduled data backup time 764 until the usable capacity of the target instance becomes larger than the target data size. The data of the selected instance in the high-speed storage device 221 is rearranged in the low-speed storage device 222, and the process proceeds to S360 described above. Thereby, the hierarchy control unit 311 can arrange the target data of the target instance in the high-speed storage device 221.

After the above-described primary site data placement processing, the primary site host server 100 performs data backup of the target instance, so that the storage device 200 at the primary site can quickly write data written to the target instance since the previous data backup. Data can be written to the storage device 221. Further, when the usable capacity of the high-speed storage device 221 is insufficient, the primary site storage apparatus 200 can preferentially write data to be backed up next to the high-speed storage device 221.

In S 170 described above, the secondary site storage apparatus 200 performs secondary site data placement processing in response to the reception of the statistical information table 760 from the primary site management server 400.

FIG. 12 shows the secondary site data placement process.

In S510, the hierarchy control unit 311 of the storage device 200 at the secondary site acquires and stores the statistical information table 760 from the management server 400 at the primary site.

In S520, the hierarchy control unit 311 calculates the load target data size, which is the total size of the load target data, for each SLA based on the statistical information table 760. The load target data is data accessed (Read or Write) within the monitoring period among the load target instances. The load target instance is, for example, an instance having an SLA equal to or higher than a specific SLA.

In S530, the hierarchy control unit 311 determines whether or not the total capacity of the high-speed storage device 221 is equal to or greater than the sum of the load target data sizes of all SLA.

When the result of S530 is Yes, in S540, the hierarchy control unit 311 arranges the load target data of all the SLAs in the high speed storage device 221 and arranges data other than the load target data in the low speed storage device 222. finish.

If the result of S530 is No, in S550, the hierarchy control unit 311 selects data having the total capacity of the high-speed storage device 221 in order of priority from the load target data, and stores the selected data at high speed. It arrange | positions to the device 221, and arranges other data to the low-speed storage device 222, and complete | finishes this flow. Here, the hierarchy control unit 311 may determine the SLA of the instance as the priority of the instance. As a result, the storage device 200 at the secondary site can place the data of the instance having a high SLA in the high speed storage device 221 with priority. Further, the hierarchy control unit 311 may determine the classification of the data access frequency 767 as the priority of the data. Further, the hierarchy control unit 311 may determine the priority of data based on the SLA and the access frequency 767. Thereby, the storage device 200 at the secondary site can preferentially arrange the data in the area having a high access frequency to the high-speed storage device 221.

After the above secondary site data placement processing, the primary site storage apparatus 200 and the secondary site storage apparatus 200 perform replication, so that the data accessed after the previous data backup in the primary site host server 100 is stored in the secondary site. You can write to the high-speed storage device 221 at your site. As a result, the host server 100 at the secondary site loads data from the high-speed storage device 221 at the time of failover, so that recovery can be speeded up and RTO can be shortened. Further, when the capacity of the high-speed storage device 221 is insufficient, the storage device 200 at the secondary site can place the data having a high priority among the data accessed within the monitoring period with priority on the high-speed storage device 221. it can.

The first memory corresponds to the memory 110 etc. in the primary site 10. The first host processor corresponds to the processor 120 or the like in the primary site 10. The first upper storage device corresponds to the high-speed storage device 221 or the like in the primary site 10. The first lower level storage device corresponds to the low speed storage device 222 or the like in the primary site 10. The first storage processor corresponds to the processor 320 or the like in the primary site 10. The first management processor corresponds to the processor 420 and the like in the primary site 10. The first in-memory database corresponds to the in-memory database in the primary site 10 or the like. The second memory corresponds to the memory 110 or the like in the secondary site 20. The second host processor corresponds to the processor 120 or the like in the secondary site 20. The second upper storage device corresponds to the high-speed storage device 221 or the like in the secondary site 20. The second lower level storage device corresponds to the low speed storage device 222 or the like in the secondary site 20. The second storage processor corresponds to the processor 320 and the like in the secondary site 20. The second management processor corresponds to the processor 420 and the like in the secondary site 20. The second in-memory database corresponds to the in-memory database in the secondary site 20 or the like. The access log corresponds to the access log table 710 and the like. The access information corresponds to the statistical information table 760, the statistical information 540, and the like. The backup corresponds to a data backup or the like.

As mentioned above, although embodiment of this invention was described, this is an illustration for description of this invention, Comprising: It is not the meaning which limits the scope of the present invention to the said structure. The present invention can be implemented in various other forms.

10: Primary site, 20: Secondary site, 100: Host server, 110: Memory, 120: Processor, 200: Storage device, 210: Controller, 220: Recording unit, 221: High-speed storage device, 222: Low-speed storage device, 310 ... Memory, 311 ... Hierarchy control unit, 320 ... Processor, 400 ... Management server, 410 ... Memory, 420 ... Processor

Claims

A first memory;
A first host processor connected to the first memory;
A first upper storage device;
A first lower storage device having a performance lower than that of the first upper storage device;
A first storage processor connected to the first host processor, the first upper storage device, and the first lower storage device;
A first management processor connected to the first host processor and the first storage processor;
With
The first host processor is
By executing the first in-memory database, the data of the first in-memory database is written to the first memory;
Recording an access log of the first memory;
The first management processor includes:
Based on the access log, generating access information indicating data accessed in the first memory by the first in-memory database within a specific period;
The first storage processor is
Controlling a first storage tier using the first upper storage device and the first lower storage device;
Based on the access information, data written to the first memory within the specific period is arranged in the first upper storage device.
Computer system.
A second memory;
A second host processor connected to the second memory;
A second upper storage device;
A second lower storage device having a performance lower than that of the second upper storage device;
A second storage processor connected to the second host processor, the second upper storage device, and the second lower storage device;
Further comprising
The second storage processor
Controlling a second storage tier using the second upper storage device and the second lower storage device;
Based on the access information, data accessed within the specific period in the first memory is arranged in the second upper storage device.
The computer system according to claim 1.
When a failure occurs in the first in-memory database, the second host processor copies the data stored in the second upper storage device to the second memory and uses the data in the second memory. Execute a second in-memory database;
The computer system according to claim 2.
The first storage processor arranges data in the first in-memory database other than data arranged in the first upper storage device in the first lower storage device;
The computer system according to claim 3.
The second storage processor arranges data in the first in-memory database other than data arranged in the second upper storage device in the second lower storage device;
The computer system according to claim 4.
The first host processor periodically performs backup to write data written to the first memory within the specific period to the first storage tier,
The specific period is a period between two consecutive backups.
The computer system according to claim 5.
The first host processor executes a plurality of instances of the first in-memory database, periodically performs backup to write data written to the first memory for each instance to the first storage tier,
The first storage processor places the data of the next backup instance in the first upper storage device in preference to other instances.
The computer system according to claim 6.
The first storage processor is
Select a target instance from the plurality of instances,
Based on the access information, the usable capacity of the first upper storage device by the target instance is calculated,
Calculate the size of the target data for backup of the target instance,
When the target instance is the next instance to be backed up and the usable capacity is smaller than the size of the target data, the instances other than the target instance among the data stored in the first upper storage device To increase the usable capacity by placing the data in the first lower storage device,
The computer system according to claim 7.
The second storage processor
Based on the access information, the priority of a plurality of data accessed within the specific period in the first memory is determined,
When the capacity of the second upper storage device is smaller than the size of the plurality of data, the data having a higher priority among the plurality of data is preferentially arranged in the second upper storage device;
The computer system according to claim 8.
The access information includes SLA (service level agreement) set in advance for each instance,
The second storage processor determines data priority of each instance based on the SLA among the plurality of data.
The computer system according to claim 9.
The first host processor includes the access frequency of each of the plurality of data in the access information,
The second storage processor determines the priority of each data based on the access frequency;
The computer system according to claim 10.
By executing the first in-memory database, the data of the first in-memory database is written to the first memory;
Recording an access log of the first memory;
Based on the access log, generating access information indicating data accessed in the first memory by the first in-memory database within a specific period;
Controlling a first storage tier using a first upper storage device and a first lower storage device having a performance lower than that of the first upper storage device;
Based on the access information, data written to the first memory within the specific period is arranged in the first upper storage device.
To be prepared,
Data storage method.