WO2015029208A1 - データベース管理装置、データベース管理方法及び記憶媒体 - Google Patents
データベース管理装置、データベース管理方法及び記憶媒体 Download PDFInfo
- Publication number
- WO2015029208A1 WO2015029208A1 PCT/JP2013/073271 JP2013073271W WO2015029208A1 WO 2015029208 A1 WO2015029208 A1 WO 2015029208A1 JP 2013073271 W JP2013073271 W JP 2013073271W WO 2015029208 A1 WO2015029208 A1 WO 2015029208A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- task
- resource
- query
- database
- database management
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
- G06F16/24534—Query rewriting; Transformation
- G06F16/24542—Plan optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
- G06F16/24534—Query rewriting; Transformation
- G06F16/24542—Plan optimisation
- G06F16/24545—Selectivity estimation or determination
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/21—Design, administration or maintenance of databases
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2455—Query execution
Definitions
- the present invention relates to data management technology.
- DB database
- DBMS database management system
- Patent Document 1 A technique disclosed in Patent Document 1 is known as a technique for shortening a data read waiting time in processing of one query.
- the DBMS generates a plan (hereinafter referred to as a query execution plan) in which a plurality of database operations (DB operations or processing steps) necessary for executing a query are combined.
- a task for executing a step is dynamically generated, and data read requests are multiplexed by executing the tasks in parallel.
- the DBMS increases the task up to a predetermined number, and then generates a task so as to maintain the predetermined number.
- processor cores a processor equipped with a plurality of cores that perform operations
- a computer has a plurality of processor cores by mounting a plurality of such processors.
- a storage device has been directly connected to a hard disk drive (HDD).
- HDD hard disk drive
- a storage device having a plurality of HDDs connected thereto via a high-speed network is also common.
- a DBMS that dynamically generates tasks uses multiple tasks to execute resources such as computer CPU resources, I / O resources, memory resources, etc. It is shortened.
- I / O Input / Output
- An I / O request is an input request from an external device or an output request to an external device, and is issued from a computer executing a DBMS to the external device, for example.
- the I / O request issued to the storage apparatus becomes an input request from the storage apparatus or an output request to the storage apparatus.
- An I / O request issued to another computer is an input request from another computer or an output request to another computer.
- the conventional DBMS changes in processing for each query, and uses CPU resources, memory resources, I / O resources, and the like. For this reason, there are cases where CPU resources, memory resources, or I / O resources cannot be used sufficiently even if the number of tasks is set appropriately.
- an object of the present invention is to control task generation so that a DBMS that dynamically generates tasks can effectively use resources such as CPU resources, memory resources, and I / O resources.
- the CPU resource means a CPU (processor core) or a processing capability of the CPU, and is consumed by execution processing of a task to which a DB operation by the DBMS is assigned.
- the memory resource means the memory (free storage area) installed in the computer and is consumed when the task stores data in the memory.
- An I / O resource refers to a resource that affects the performance of I / O requests (data read waiting time, etc.) issued by executing a task in a DBMS.
- a computer, a storage device, or a computer The performance of the network or the like connecting the storage or the computer, the storage device, or the network corresponds to the I / O resource. Specifically, it is the performance of the network adapter for the computer to connect to the network, and the performance (bandwidth, etc.) of the network cable that transfers the I / O request and data corresponding to the I / O request.
- the performance of the network adapter for the storage apparatus to connect to the network in order to process the I / O request issued from the computer by the storage apparatus, or for processing the I / O request by the storage apparatus The processing capacity of the processor, the capacity of the HDD for recording data, and the like correspond to I / O resources.
- I / O resources the number of I / O requests that can be issued simultaneously by the network adapter, the upper limit of the data transfer rate of the network cable, the number of HDDs that process I / O requests, etc. Affects O performance. These I / O resources are consumed when processing an I / O request by executing a task in the DBMS.
- the present invention is a database management apparatus that manages a database stored in a storage apparatus, and includes a query reception unit that receives a query to the database, and information that represents one or more database operations necessary to execute the received query
- a query execution plan generation unit that generates a query execution plan that includes, and dynamically executes a task to execute a database operation when executing a query received based on the generated query execution plan.
- a query execution unit that executes the generated task the query execution unit obtains the usage status of the resources used to execute the accepted query, and follows the database operation executed by the generated task.
- resource usage Generates a new task on the basis of, it is to execute a new task in parallel with other tasks.
- the task in a database management apparatus that dynamically generates a task, the task is generated so that the resources used for executing the query such as a CPU resource, a memory resource, or an I / O resource are sufficient. It becomes possible.
- the database management device when the database management device generates a task, it generates a task that uses an insufficiently used resource, and is used for executing a query such as a CPU resource, a memory resource, or an I / O resource. Resource utilization can be improved.
- Example 1 is a block diagram illustrating an example of a computer system according to a first embodiment of this invention. It is Example 1 of this invention and is a block diagram which shows an example of DBMS. It is a figure which shows Example 1 of this invention and demonstrates the definition of the table
- FIG. 5 is a diagram illustrating an example of the arrangement of the database in the storage areas 1 to 4 according to the first embodiment of this invention. It is a figure which shows Example 1 of this invention and shows an example of a DB area management table. It is a figure which shows Example 1 of this invention and shows an example of a query execution plan. It is a figure which shows Example 1 of this invention and shows an example of the cost table which set CPU cost in each process step.
- Example 1 of this invention shows an example of task execution state information. It is a figure which shows Example 1 of this invention and shows the 1st example of process step execution state information. It is a figure which shows Example 1 of this invention and shows the 2nd example of process step execution state information. It is a figure which shows Example 1 of this invention and shows the 3rd example of process step execution state information. It is a figure which shows Example 1 of this invention and shows an example of the data structure of task management information. It is a figure which shows Example 1 of this invention and shows an example of a context. It is a figure which shows Example 1 of this invention and shows an example of the produced
- Example 1 of this invention shows an example of the produced
- 10 is a flowchart of query execution processing according to the first embodiment of this invention.
- 5 is a flowchart illustrating task execution processing according to the first embodiment of this invention.
- 5 is a flowchart of processing step execution processing according to the first embodiment of this invention.
- 10 is a flowchart of context generation processing according to the first embodiment of this invention.
- 10 is a flowchart of DB page acquisition processing according to the first embodiment of this invention. It is a figure which shows Example 1 of this invention and shows an example of a system performance threshold value table
- Example 1 of this invention shows an example of a performance data table.
- 10 is a flowchart of task generation processing according to the first embodiment of this invention. It is a figure which shows Example 1 of this invention and shows an example of a storage area performance data table. 10 is a flowchart of context acquisition processing according to the first embodiment of this invention. It is a figure which shows Example 1 of this invention and shows an example of a storage area performance threshold value table. 10 is a flowchart of context acquisition processing according to the second embodiment of this invention. It is a figure which shows Example 3 of this invention and shows an example of the 2nd query of a database. It is a figure which shows Example 3 of this invention and demonstrates an example of a query execution plan.
- FIG. 10 is a flowchart of context acquisition processing according to the third embodiment of this invention.
- 10 is a flowchart of context acquisition processing according to the fourth embodiment of this invention.
- 10 is a flowchart of context acquisition processing according to the fifth embodiment of this invention.
- FIG. 10 is a block diagram illustrating an example of a computer system according to a sixth embodiment of this invention.
- FIG. 10 is a diagram illustrating an example of a DB area management table according to the sixth embodiment of this invention. It is a figure which shows Example 6 of this invention and shows an example of an 8th context. It is a figure which shows Example 6 of this invention and shows an example of a 9th context. 10 is a flowchart illustrating task execution processing according to the sixth embodiment of this invention. It is Example 6 of this invention, and is a flowchart of a context acquisition process.
- FIG. 10 is a block diagram illustrating an example of a computer system according to a seventh embodiment of this invention.
- FIG. 1A is a block diagram illustrating an example of a computer system according to Embodiment 1 of the present invention.
- FIG. 1B is a block diagram illustrating an example of a DBMS executed by the computer 100.
- the computer system includes a computer 100 and an external storage device 200.
- the computer 100 and the external storage apparatus 200 are connected via a communication network 300.
- a protocol for communication via the communication network 300 for example, FC (Fibre Channel), SCSI (Small Computer System Interface), IB (Infini Band), or TCP / IP (Transmission Control Protocol / Internet Protocol) is adopted. can do.
- the computer 100 is, for example, a personal computer, a workstation, or a main frame.
- the computer 100 includes a network adapter 110, a processor (typically a microprocessor (for example, a CPU (Central Processing Unit))) 120, a local storage device 130, and a memory 140.
- a processor typically a microprocessor (for example, a CPU (Central Processing Unit))
- a local storage device 130 for example, a hard disk drive, or a main frame.
- a memory 140 typically a microprocessor (for example, a CPU (Central Processing Unit)
- the processor 120 executes a computer program, for example, an OS (Operating System) (not shown) or a DBMS (Data Base Management System) 141.
- the one or more processors 120 have one or more processor cores. Each processor core can execute processing independently or in parallel.
- the memory 140 temporarily stores a program executed by the processor 120 and data used by the program.
- the memory 140 stores a DBMS 141 that is a program for managing a DB (DataBase) and a series of related processes and data.
- the memory 141 may store an AP (Application Program) 148 for issuing a query to the DBMS 141.
- the local storage device 130 stores the program of the computer 100 and data used by the program.
- the local storage device 130 is a nonvolatile storage medium, for example, a magnetic disk, a flash memory, or other nonvolatile semiconductor memory.
- the network adapter 110 connects the communication network 300 and the computer 100.
- the processor 120 may be an element included in a control device connected to the network adapter 110, the memory 140, and the like.
- the control device may include dedicated hardware circuits (for example, a circuit that performs encryption and / or decryption of data and a circuit that compresses and / or decompresses data).
- the computer 100 may include a plurality of at least one of the network adapter 110, the processor 120, the local storage device 130, and the memory 140 from the viewpoint of performance and redundancy.
- the computer 100 may include an input device (for example, a keyboard and a pointing device) and a display device (for example, a liquid crystal display) not shown.
- the input device and the display device may be integrated.
- the computer 100 is a database management apparatus in which the DBMS 141 executes a query issued to the database 206.
- This query is issued by an AP 148 executed by the computer 100 or an AP executed by a computer (client) (not shown) connected to the communication network 300.
- the DBMS 141 executes a query issued by the AP 148, and transmits an I / O request for the DB 206 stored in the external storage apparatus 200 to the external storage apparatus 200 via the OS in accordance with the execution of the query.
- DBMS 141 only one DBMS 141 is executed by the computer 100, but a plurality of DBMSs 141 may be executed.
- the OS (not shown) may be a guest OS executed on a virtual machine generated by the virtualization program. Then, the OS on the virtual machine may execute the DBMS 141. There may be a plurality of virtual machines executed by the computer 100.
- External storage device 200 stores data used by computer 100.
- the external storage apparatus 200 receives an I / O request from the computer 100, executes processing corresponding to the I / O request, and transmits the processing result to the computer 100.
- the external storage apparatus 200 includes a network adapter 201, a storage device group 203, and a controller 202 connected to them.
- the network adapter 201 connects the external storage device 200 to the communication network 300.
- the storage device group 203 includes one or more storage devices.
- the storage device is a non-volatile storage medium, such as a magnetic disk, a flash memory, or other non-volatile semiconductor memory.
- the storage device group 203 may be a group that stores data at a predetermined RAID level according to RAID (Redundant ARRAY of Independent Disks).
- a logical storage device (logical volume) based on the storage space of the storage device group 203 may be provided to the computer 100.
- the storage device group 203 stores the DB 206.
- the DB 206 includes one or more tables 205 and indexes 204.
- Table 205 is a set of one or more records, and each record is composed of one or more columns.
- the index 204 is a data structure generated for one or more columns in the table 205, and speeds up the access to the table 205 by a selection condition including the column targeted by the index.
- the index is a data structure that holds information (RowID) for specifying a record in a table including the value for each value of a target column, and a B-tree structure or the like is used.
- RowID information
- An example of the configuration of the DB table 205 and an example of the relationship between the tables 205 will be described later.
- the controller 202 includes, for example, a memory and a processor (not shown), and executes data input / output with respect to the storage device group 203 storing the DB 206 in accordance with an I / O request from the computer 100.
- the controller 202 stores data to be written in accordance with a write request from the computer 100 in the storage device group 203, reads data to be read in accordance with a read request from the computer 100 from the storage device group 203, and stores the data. It transmits to the computer 100.
- the external storage apparatus 200 may include a plurality of elements such as the controller 202 from the viewpoint of ensuring performance and ensuring redundancy.
- a plurality of external storage devices 200 may be provided.
- the DBMS 141 manages the DB 206 including business data. As shown in FIG. 1B, the DBMS 141 includes a client communication control unit 142, a query execution plan generation unit 143, a query execution unit 144, an execution task management unit 145, a DB buffer management unit 146, a DB buffer 1460, a cost table 1431, and a DB area management. Table 147 is included.
- the client communication control unit 142 controls communication between the client connected to the communication network 300 or the AP 148 executed by the computer 100. Specifically, the client communication control unit (query receiving unit) 142 receives a query issued from a client (not shown) or the AP 148, and executes a process of responding to the client or the AP 148 with a query processing result.
- the query is described in, for example, SQL (Structured Query Language).
- the query execution plan generation unit 143 generates a query execution plan having one or more processing steps necessary for executing the query received by the client communication control unit 142.
- the query execution plan is, for example, information in which the execution order of processing steps to be performed at the time of executing a query is defined in a tree structure, and is stored in the memory 140. An example of the query execution plan will be described later.
- the DB buffer management unit 146 manages a DB buffer 1460 (or DB cache) as a storage area for temporarily storing data in the DB 206.
- the DB buffer 1460 is constructed on the memory 140.
- the DB buffer may be constructed on the local storage device 130.
- the query execution unit 144 dynamically generates a task for processing according to the query execution plan generated by the query execution plan generation unit 143, and accesses the DB 206 by executing the above-described task, and generates a query result.
- the query execution unit 144 returns the access result to the DB 206 generated by the task to the query issuer.
- the query execution unit 144 includes a task generation control unit 152, a context management unit 153, a system performance threshold table 154, and a performance data table 155.
- the task generation control unit 152 generates a new task based on the resource usage status when a task generation request is received.
- the resources are CPU resources (processor 120), I / O resources (network adapter 110, storage device group 203, and the like), memory resources (memory 140), and the like of computers used for executing queries.
- the usage status of the CPU resource is indicated by the CPU usage rate.
- the usage status of the I / O resource indicates the amount of data transferred from an external device such as the external storage device 200 to the computer 100, the amount of data transferred from the computer 100 to an external device such as the external storage device 200, or the computer 100 externally. This is indicated by the number of I / O requests issued to an external device such as the storage 200.
- the data transfer amount may be a data transfer rate that is a data transfer amount per unit time, or may be a cumulative data transfer amount that is an accumulated data transfer amount.
- the number of I / O requests may be IOPS (Input / Output Per Second) that is the number of I / O requests processed per unit time, or may be the cumulative number of I / O requests that is the cumulative number of I / O requests. It may be the number of outstanding I / Os, which is the number of issued I / O requests for which I / O has not been completed.
- the usage status of the memory resource is indicated by the memory usage. For example, the task generation control unit 152 generates a task when the use of CPU resources or I / O resources is insufficient.
- the task generation control unit 152 acquires the current CPU usage rate, the disk transfer rate (or data transfer rate) of the external storage device 200 or the local storage device 130, and the usage status of resources such as IOPS. Compared with a preset CPU utilization threshold, disk transfer speed threshold, and IOPS threshold. If the usage amount is less than the threshold, the task generation control unit 152 determines that resources such as CPU resources and I / O resources are not sufficiently used (or does not satisfy a predetermined condition).
- the task generation control unit 152 generates a task if the resource usage status is insufficient. In order to determine whether the resource usage is sufficient, the task generation control unit 152 refers to the value from the system performance threshold table 154, and the current CPU usage rate, disk transfer rate, and IOPS are the performance data table 155. Refer to
- the context management unit 153 manages a context including information necessary for executing a newly generated task.
- the context includes a first information indicating which one of one or more processing steps represented by the query execution plan is a processing step that starts execution in a task, and a processing step indicated by the first information.
- This is information including second information relating to an access destination of necessary data and third information relating to data necessary for generating a result by a task.
- the structure of context management information which is information for managing the context, will be described later.
- the system performance threshold value table 154 holds threshold values for determining whether or not the CPU resource and the I / O resource are sufficiently used, and the task generation control unit 152 refers to the threshold value.
- the performance data table 155 holds values for determining the current usage status of CPU resources and I / O resources, and is referenced by the task generation control unit 152.
- the execution task management unit 145 manages a task for executing a query.
- the task may be a process or thread managed by the OS 415, or a pseudo process or pseudo thread implemented by the DBMS 412.
- the task may be a set of pointers (function pointers) to a function in which each process is collected as a function.
- the structure of task management information which is information for managing tasks, will be described later.
- At least a part of processing performed by at least one processing unit of the client communication control unit 142, the query execution plan generation unit 143, the query execution unit 144, and the DB buffer management unit 146 may be performed by hardware.
- the processing unit when the processing unit is the subject, the processing is actually performed by the processor 120 that executes the processing unit, but at least a part of the processing unit is realized by hardware.
- the hardware may be the subject instead of or in addition to the processor 120.
- a computer program for realizing the DBMS 141 may be installed in the computer 100 from a program source.
- the program source may be, for example, a storage medium readable by the computer 100 or another computer.
- the configuration of the DBMS 141 shown in FIG. 1 is an example.
- a certain processing unit may be divided into a plurality of processing units, or a single processing unit in which functions of a plurality of processing units are integrated may be constructed.
- the function units of the client communication control unit 142, the query execution plan generation unit 143, the query execution unit 144, the execution task management unit 145, and the DB buffer management unit 146 constituting the DBMS 141 are loaded into the memory 140 as programs, and the processor 120.
- the processor 120 operates as a functional unit that realizes a predetermined function by executing processing according to a program of each functional unit.
- the processor 120 functions as the query execution unit 144 by executing processing according to the query execution program.
- the processor 120 also operates as a functional unit that implements each of a plurality of processes executed by each program.
- a computer and a computer system are an apparatus and a system including these functional units.
- Information such as programs and tables for realizing each function of the DBMS 141 is stored in a local storage device 130, an external storage device 200, a nonvolatile semiconductor memory, a hard disk drive, a storage device such as an SSD (Solid State Drive), or an IC card, SD. It can be stored in a computer-readable non-transitory data storage medium such as a card or DVD.
- FIG. 2 is a diagram illustrating the definitions of the table 205 and the index 204 of the DB 206 according to the first embodiment.
- the DB 206 includes, as the table 205, for example, a Part table 2051 (see FIG. 3) including the columns c1 (2052) and c2 (2053), and a Lineitem table 2054 (see FIG. 3) including the columns c3 (2055) and c4 (2056). 4).
- the DB 206 uses, as the index 204, an index (Part index) related to the Part table 2051 based on the value of the column c1 (2052) and an index (Lineitem index) related to the Lineitem table based on the value of the column c3 (2055). Have.
- FIG. 3 is a diagram illustrating an example of the Part table 2051 of the DB 206 according to the first embodiment.
- the Part table 2051 of the DB 206 is a table that logically associates the value of the column c1 (2052) with the value of the corresponding column c2 (2053), for example.
- FIG. 4 is a diagram illustrating an example of the Lineitem table 2054 of the DB 206 according to the first embodiment.
- the Lineitem table (2054) of the DB 206 is a table in which, for example, the value of the column c3 (2055) is associated with the value of the corresponding column c4 (2056).
- FIG. 5 is a diagram illustrating an example of the first query of the DB according to the first embodiment.
- the query shown in FIG. 5 is an example of a query for the DB 206 having the structure shown in FIGS. 2 to 4 and 6 to 8.
- the value of the column c1 (2052) is “130” and the value of the column c2 (2053) and the value of the column c3 (2055) are the same from the Part table 2051 and the Lineitem table 2054. For some, this means extracting the value of column c1 (2052) and the value of column c4 (2056).
- FIG. 6 is a diagram illustrating an example of the data structure of the part index 2041 in the DB 206 according to the first embodiment.
- the Part index 2041 is, for example, a B-tree structure for searching for a page (P, P1 to P9) of a part table storing a corresponding record and a slot in the page based on the value of the column c1 (2052).
- the data structure of the index may be a data structure such as an R-tree, hash, or bitmap.
- the Lineitem index 2042 has a B-tree structure for searching, for example, the page of the Lineitem table 2054 storing the corresponding record and the slot in the page based on the value of the column c3 (2055).
- a page is a minimum data unit in input / output with respect to the DB 206.
- the Part index 2041 has a B-tree structure with the page P as an input / output unit.
- leaf pages P4 to P9 which are the lowest pages and upper pages P1 to P3 which are upper pages of the leaf pages.
- the highest page (P1) among the upper pages P1 to P3 is referred to as a root page.
- One or more associated entries are provided.
- the page P1 has a pointer Pt12 to the page P2 that manages the correspondence to the value of the column c1 (2052) of “100” or less, and the correspondence to the value of the column c1 that is larger than “100” and “200” or less. Is stored, a pointer Pt13 to the page P3 that manages.
- the pointer to the page in the lower hierarchy of each page is associated with the maximum value of the value of the column c1 (2052) managed in the page in the lower hierarchy.
- One or more entries Pt are provided.
- the value of the column c1 (2052) and the RowID that is the storage position of the record of the Part table corresponding to the value (for example, the page number of the Part table 2051 and the slot number in the page).
- One or more associated entries E are stored.
- the page P7 which is a leaf page, includes an entry E71 including the page and slot number in which the value of the column c2 (2053) corresponding to the value “110” of the column c1 (2052) is stored, and the column c1 ( 2052) and the entry (E72) including the number of the page and slot in which the value of the column c2 (2053) corresponding to the value “130” is stored.
- the other leaf pages are the same, and are represented by entry E in the figure.
- FIG. 7 is a diagram illustrating an example of a data structure that holds a RowID in which the key value of the Part index 2041 in the DB 206 according to the first embodiment is 130.
- the RowID (20411) of the record corresponding to the value “130” of the column c1 (2052) of the Part table 2051 shown in FIG. 3 includes slot 2 of page P21, slot 1 of page P22, and slot 4 of page P23. In total, 10 RowIDs are stored.
- FIG. 8 is a diagram for explaining an example of the data structure of the Part table (2051) according to the first embodiment.
- the record in slot 2 of page P21 has column c1 of 130 and column c2 of id 131.
- the column c1 is 130 and the column c2 is id132.
- the column c1 is 130 and the column c2 is id133.
- the RowID of the record corresponding to the value “130” of the column c1 illustrated in FIG. 7 indicates the ten records 20411 in which the column c1 is 130.
- FIG. 9 is a block diagram illustrating an example in which the DB 206 according to the first embodiment is arranged in a plurality of storage areas of the external storage apparatus 200.
- the external storage device 200 is composed of a plurality of logical storage areas (logical volumes or logical units: LU) # 1 to # 4.
- the index 204 and the table 205 of the DB 206 are stored in these storage areas # 1 to # 4.
- An example of distributed arrangement is shown.
- the index 204 is divided into two parts, a Part index 2041 and a Lineitem index 2042.
- the Lineitem table 2054 is divided into four parts: LINEITEM (1) 2054-1 to LINEITEM (4) 2054-4.
- the Part index IDX_PART 2041 is stored in the storage area # 1, and the Lineitem index IDX_LINEITEM 2041 is stored in the storage area # 2.
- the Part table 2051 includes four areas (PART (1) 2051-1, PART (2) 2051-2, PART (3) 2051-3, and PART (4) 2051-4).
- PART (1) 2051-1 is in the storage area # 1
- PART (2) 2051-2 is in the storage area # 2
- PART (3) 2051-3 is in the storage area # 3
- PART (4) 2051. -4 is stored in the storage area # 4.
- the Lineitem table 2054 includes four areas 2054-1 to 2054-4, and the areas 2054-1 to 2054-4 are stored in the storage areas # 1 to # 4.
- FIG. 10 is a diagram illustrating an example of the DB area management table 147 according to the first embodiment.
- the DB area management table 147 in FIG. 10 stores a DB object (Part index, Part table, Lineitem index, Lineitem table in this embodiment) 1471, a page number 1472 constituting the DB object, and any page number stored therein.
- DB object 1471 is the Part index IDX_PART
- the page numbers 1472 are P1 to P20
- the page numbers P1 to P20 are stored in the storage area # 1.
- the page numbers of PART (2) constituting the Part table 2051 are P121 to P150, and indicate that the pages P121 to P150 are stored in the storage area # 2.
- the DBMS 141 holds the DB area management table 147, but it may be held in the storage device group 203 of the external storage apparatus 200.
- FIG. 11 is a diagram illustrating an example of a query execution plan according to the first embodiment.
- the query execution plan shown in FIG. 11 shows an example of a query execution plan generated by the query execution plan generation unit 143 when the DBMS 141 accepts the query shown in FIG.
- the query execution plan corresponding to the query shown in FIG. 5 includes processing step # 1 for performing an index search using the part index 2041, processing step # 2 for obtaining a record from the part table 2051, and a line item index.
- the outer side in the figure indicates a loop outside the nest. In this embodiment, the processing is performed on the Part table 2051. Further, the inner side in the figure indicates a loop inside the nest, and in this embodiment, the processing is performed on the Lineitem table 2054.
- FIG. 12 is a diagram showing an example of a cost table 1431 in which CPU costs are set in the processing steps # 1 to # 6 shown in FIG.
- the cost table 1431 is managed on the memory 140 by the query execution plan generation unit 143.
- the cost table 1431 includes one entry including a processing step 1432 for storing the number of the processing step and a CPU cost 1433 for storing the cost of the processor 120.
- the CPU cost 1433 in each processing step # 1 to # 6 is a numerical value representing the processing amount required by the processor 120 when executing the corresponding processing step.
- the processing amount for example, the number of instructions in each processing step, the number of clocks required by the processor 120 to execute each processing step, the processing time required for processing in the processing step, or the like can be employed.
- the CPU cost 1433 is represented by the processing time ( ⁇ sec) when each processing step 1432 is executed by the computer 100.
- the query execution plan generation unit 143 for example, the query execution plan is determined based on the CPU cost and I / O cost of each processing step. Therefore, the CPU cost is the CPU cost used in the cost estimation of the query execution plan generation unit 143. May be used.
- the query execution plan generation unit may set the result of measuring the processing time for each processing step and the number of clocks required for processing in the cost table 1431 in advance 143. Further, the CPU cost may be detected during execution of the query, and this value may be set in the cost table 1431. Furthermore, the detected CPU cost may be corrected using the processing time of the processing step during query execution.
- FIG. 13 is a diagram illustrating an example of the task execution state information 73 according to the first embodiment.
- the task execution state information 73 is held on the memory 14 by the execution task management unit 145.
- the task execution state information 73 includes a work area 73a set on the memory 140, a processing step 73b for storing the number of the processing step to be executed, and a processing step execution state 73c.
- the work area 73a stores a pointer indicating the work area 73d for storing a column value when the corresponding task processes the query execution plan.
- the processing step 73b stores information for identifying a processing step to be executed by the corresponding task, for example, a processing step number.
- the processing step execution state 73c stores execution state information (processing step execution state information) 74 of the corresponding processing step. A specific example of the processing step execution state information 74 will be described later.
- FIG. 14 is a diagram illustrating a first example of the execution state information 74 of the processing steps according to the first embodiment.
- FIG. 14 shows processing step execution state information for a task that uses the upper page in the index search.
- the processing step execution state information 74A includes a search condition 74a, a page number 74b, and a slot number 74c.
- the search condition is stored in the search condition 74a.
- the page number 74b stores the number of the upper page (page number) used in task processing.
- the slot number 74c stores the slot number (slot number) in the page used in the task processing.
- FIG. 15 is a diagram illustrating a second example of the execution state information 74 of the processing steps according to the first embodiment.
- FIG. 15 shows processing step execution state information for a task that uses leaf pages P4 to P9 (see FIG. 6) in the index search.
- the process step execution state information 74B includes a search condition 74d, a page number 74e, a slot number 74f, and a process Row ID number 74g.
- the search condition is stored in the search condition 74d.
- the page number 74e stores the page number “7” of the leaf page used in task processing.
- the slot number 74f stores the slot number “2” of the slot in the page used for task processing.
- the processing Row ID number 74g stores the ID number of the Row (processing Row ID number) “1” in the slot processed by the corresponding task.
- FIG. 16 is a diagram illustrating a third example of the execution state information 74 of the processing steps according to the first embodiment.
- FIG. 16 shows process step execution state information 74C for a task that acquires a record.
- the processing step execution state information 74C includes a page number 74h and a slot number 74i.
- the page number 74h stores the page number “2” of the page used in task processing.
- the slot number 74i stores the slot number “2” of the slot in the page used for task processing.
- FIG. 17 is a diagram illustrating an example of a data structure of the task management information 1450 according to the first embodiment.
- the task management information 1450 has a data structure managed by the execution task management unit 145 and is held on the memory 140.
- the data structure of the task management information 1450 includes an executable list 1451 for managing executable tasks and a waiting list for managing tasks that are waiting to be executed, such as tasks waiting for completion of I / O requests. 1452 is included.
- the executable list 1451 has a pointer 1453 to task execution status information 73 (see FIG. 13), which is execution status information relating to executable tasks.
- the waiting list 1452 has a pointer 1454 to task execution state information 73 (see FIG. 13), which is execution state information relating to a task that is waiting.
- the task execution state information 73 includes a pointer to the task execution state information 73 related to other tasks that can be executed.
- FIG. 18 is a diagram illustrating an example of the initial context 1530.
- the context 1530 is information necessary for executing a newly generated task, and is managed by the context management unit 153.
- the initial context 1530 includes a start step 1531, an intermediate result 1532, a generatable number 1533, and an execution state 1534.
- the start step 1531 stores the number or identifier of the processing step to be executed.
- the intermediate result 1532 a pointer indicating the location of the work area 1539 for storing the intermediate result necessary for the task executing the processing step is stored.
- the intermediate result is acquired data necessary for generating a query result.
- the work area 1539 is an area set on the memory 140.
- the execution state 1534 information for specifying the content of processing of a task to be executed next is stored.
- the page number 1561 to be processed and an unprocessed data list 1562 composed of a list of data that has not been processed yet.
- the unprocessed data list 1562 can be composed of, for example, a set of page numbers and slot numbers that have not been processed.
- the number of tasks that can be generated by the task generation control unit 152 (the number of tasks that can be generated) is stored in the number of tasks that can be generated 1533.
- the number of tasks that can be generated is the number of processes that are not generated as tasks among the number of processes that logically branch in the page number of the execution state of start step 1531.
- starting step 1531 stores “processing step # 1”.
- the intermediate result 1532 stores a pointer to a work area 1539 that is an area obtained by copying the contents of the work area 73d held by the task at the time of generating the context.
- the page number 1561 stores “P7”, which is the page number of the leaf page used in the processing of the task generated from the context.
- the unprocessed data list 1562 stores the remaining nine RowID 20411 data. Since there are nine tasks that can be generated from this initial context 1530, “9” is stored in the number of tasks that can be generated 1533.
- FIG. 19 is a diagram illustrating an example of the first context 1530-1 according to the first embodiment.
- contexts 1530-1 to 1530-n are generated for each feature (or type) of a generated task so that a task to be generated can be selected depending on the usage status of I / O resources and CPU resources. be able to.
- the tasks are characterized by storage areas # 1 to # 4 that issue I / O requests, I / O size that is the size of the I / O request, I / O pattern, and CPU cost.
- An I / O pattern is a feature of an address written in a plurality of I / O requests. For example, if an I / O request to an address adjacent to the address of a previously made I / O request is processed continuously, the I / O pattern is called sequential.
- the I / O pattern is called random.
- the I / O pattern as a feature of a task indicates what feature the address of those I / O requests becomes when the task is executed.
- the DBMS 141 generates contexts 1530-1 to 1530-n for each feature classified from the initial context 1530.
- the first context 1530-1 shown in FIG. 19 includes a storage area name 1535, an I / O, in addition to the start step 1531, the intermediate result 1532, the number of possible generations 1533, and the execution state 1534 shown in FIG. Size 1536, I / O pattern 1537, and CPU cost 1538 are included.
- FIG. 19 shows an example of the first context 1530-1 in which RowIDs that access the storage area # 1 are collected when the RowID 20411 (see FIG. 7) included in the initial context 1530 shown in FIG. 18 is classified. is there. Details of processing for classifying contexts to be generated based on features will be described later.
- the task generated from the context 1530-1 accesses the storage area # 1, the I / O size 1536 is 4 KB, the I / O pattern 1537 is random, and the CPU cost 1538 is 125 tasks.
- FIG. 20 is a diagram illustrating an example of the second context according to the first embodiment.
- FIG. 20 is an example of a context 1530-2 in which RowIDs that access the storage area # 2 are collected when the RowID 20411 (see FIG. 7) included in the initial context 1530 shown in FIG. 18 is classified.
- the task generated from the context 1530-2 accesses the storage area # 2, the I / O size 1536 is 4 KB, the I / O pattern 1537 is random, and the CPU cost 1538 is 125 tasks.
- the context is classified so that the use of the I / O resource and the CPU resource can be distinguished, but the classification related to the use of the memory resource may be performed.
- the memory consumption of a task that starts processing in this context may be characterized.
- FIG. 21 shows the first embodiment, and is a flowchart of the entire processing from when the DBMS 141 receives a query until it responds with a result.
- step S1 when the client communication control unit 142 receives a query from the AP 148 (step S1), the received query is passed to the query execution plan generation unit 143, and the query execution plan generation unit 143 receives the query execution plan. Generate (step S2).
- the query execution unit 144 sets the initial task number, the upper limit task number, and the lower limit task number (step S3).
- the initial task number, the upper limit task number, and the lower limit task number are set, but any one of them may be used, or the initial task number, the upper limit task number, and the lower limit task number may be arbitrarily combined.
- the initial number of tasks is the number of tasks generated by the query execution unit 144 regardless of the usage status of CPU resources and I / O resources after the start of query execution processing.
- the initial number of tasks may be specified by the user, or the DBMS 141 may automatically calculate from the hardware configuration. For example, when the communication network 300 and the computer 100 are connected by a Fiber Channel, if the number of tags, which is the number of I / O requests that can be issued simultaneously on the Fiber Channel port, is 1024, then “Fibre Channel port number ⁇ 1024”. Is set to the initial number of tasks.
- the number of simultaneous commands received by the external storage apparatus 200 is 2048, “2048” is set as the initial task number.
- “n ⁇ 32” is set as the initial number of tasks.
- the upper limit task number is the upper limit of tasks that exist on the DBMS 141 at the same time, and even if CPU resources and I / O resources are not sufficiently utilized, it is guaranteed that tasks exceeding the upper limit task number do not exist on the DBMS 141 at the same time. .
- the upper limit number of tasks may be specified by the user, or the DBMS 141 may automatically calculate from the hardware configuration. For example, when the communication network 300 and the computer 100 are connected by a Fiber Channel, if the number of tags, which is the number of I / O requests that can be issued at the same time on the Fiber Channel port, is 1024, then the “Number of Fiber Channel ports ⁇ 1024” is set. Set the maximum number of tasks.
- the external storage apparatus 200 If the number of simultaneous commands received by the external storage apparatus 200 is 2048, “2048” is set as the upper limit number of tasks.
- the external storage apparatus 200 provides a logical volume to the computer 100 using n hard disk drives (HDDs), “n ⁇ 32” is set as the upper limit task number.
- HDDs hard disk drives
- the lower limit task number is the number of tasks to be generated regardless of the usage status of CPU resources and I / O resources.
- the lower limit task number may be specified by the user, or the DBMS 141 may automatically calculate from the hardware configuration.
- the number of processor cores is set to the lower limit task number so that the processor cores can be utilized.
- the lower limit task number sets the number of HDDs as the lower limit task number so that the HDD of the external storage apparatus 200 can be utilized.
- the query execution unit 144 generates an initial context 1530 (step S4).
- the initial context 1530 is a context 1530 that generates a task that first executes a query execution plan.
- the generated initial context 1530 is registered in the context management unit 153.
- the query execution unit 144 performs query execution processing (step S5).
- the query execution unit 144 generates a new task, performs the query by executing the task, and generates a query result for the DB 203 of the external storage apparatus 200.
- the specific contents of the query execution process will be described with reference to FIG.
- the result generated by the query execution unit 144 is returned to the AP 148 from which the client communication control unit 142 has sent the query (step S6).
- the entire process ends.
- the DBMS 141 receives a query from the AP 148, generates a task corresponding to the query, executes the task, executes access to the DB 203 of the external storage apparatus 200, and generates an access result as a query processing result. .
- FIG. 22 is a flowchart illustrating an example of a query execution process according to the first embodiment. This process is a process performed in step S5 of FIG.
- the query execution unit 144 determines whether there is a context 1530 and whether there is a task (step S11).
- the presence / absence of the context 1530 determines the presence / absence of the context 1530 registered in the context management unit 153.
- the presence / absence of a task determines whether a task exists in the execution task management unit 145. If there is no context 1530 and there is no task, the query execution process ends. On the other hand, if there is a context 1530 or there is a task, the process proceeds to step S12.
- the query execution unit 144 determines whether there is an executable task (step S12).
- the execution task management unit 145 determines whether there is an executable task. If there is no executable task, the process proceeds to step S13, and if there is an executable task, the process proceeds to step S16.
- the query execution unit 144 performs task generation processing (step S13).
- the task generation process is a process of reading a context 1530 and generating a new task. Specific processing will be described later.
- the query execution unit 144 determines whether there is a task that can be executed again (step S14). If there is no task that can be executed, there is no context, and all the existing tasks are held in the pointer 1454 (see FIG. 17) of the waiting list 1452, so sleep for a certain time (step S15).
- step S12 if there is a task that can be executed in step S12 or step S14, the query execution unit 144 selects one task (step S16) and executes the selected task (step S17).
- the task execution is to start the task execution process shown in FIG. 23 if it is a new task.
- the processing is resumed from the processing when the pointer 1454 of the waiting list 1452 is entered.
- the query execution unit 144 returns to step S11 and repeats the above processing until the context 1530 and the task disappear.
- FIG. 23 is a flowchart illustrating an example of task execution processing according to the first embodiment. This process is a process performed for a new task in step S17 of FIG.
- This task execution process is applied when the query execution unit 144 executes a new task whose process has not been determined.
- the query execution unit 144 performs context acquisition processing in the context management unit 153 in order to determine the content of the new task processing (step S21). The specific contents of the context acquisition process will be described with reference to FIG.
- the query execution unit 144 sets task execution state information 73 (see FIG. 13) using the acquired context 1530 (step S22).
- the first context 1530-1 shown in FIG. 19 will be described as an example.
- the query execution unit 144 copies the value (processing step # 1) of the start step 1531 of the context 1530-1 to the processing step 73b of the task execution state information 73.
- the query execution unit 144 copies the data of the work area 1539 indicated by the pointer of the intermediate result 1532 of the context 1530 to the work area indicated by the pointer of the work area 73a of the task execution state information 73.
- the query execution unit 144 extracts one piece of data from the unprocessed data list 1562 of the acquired context 1530, and sets the processing step execution state information 74C. Since the query execution unit 144 extracts one piece of data from the unprocessed data list 1562, the number of possible generations 1533 of the context 1530 is reduced by one.
- the query execution unit 144 acquires the RowID (P22, 1) illustrated in FIG. 7 from the unprocessed data list 1562.
- the data in the unprocessed data list 1562 is a RowID for acquiring a record.
- the query execution unit 144 prepares the process step execution state information 74C of FIG. 16 which is a step execution state for acquiring a record. That is, the query execution unit 144 sets “P22” to the page number 74h of the process step execution state information 74C and sets “1” to the slot number 74i.
- the query execution unit 144 advances the process step 73b in the task execution state information 73 to set the process step # 2 as shown in FIG. .
- the query execution unit 144 completes the process of setting the task execution state information 73.
- the query execution unit 144 executes a processing step execution process according to the state set in step S22 (step S23).
- the process step execution process will be described with reference to FIG. When the process step execution process ends, the task execution process ends.
- FIG. 24 is a flowchart of processing step execution processing according to the first embodiment. This process is a process executed in step S23 of FIG.
- the query execution unit 144 acquires a page (DB page) to be queried from the DB 206 by causing the DB buffer management unit 146 to execute a DB page acquisition process (see FIG. 26) (Step S30).
- the query execution unit 144 determines whether there is data that matches the search condition for the data in the acquired page (step S31). For example, if it is an upper page of the index, it is a search process in the upper page, and if it is a leaf page, it is a search process of a leaf page. As a result of this determination, if there is no data that matches the search condition in the data on the page (“false” in step S31), the processing step execution process ends.
- step S31 when there is data that matches the search condition (“true” in step S31), the query execution unit 144 determines whether there is one data that matches the search condition, or whether there are two or more data. Determination is made (step S32).
- step S32 If the result of this determination is that there is one data that matches the search condition (“1” in step S32), the query execution unit 144 advances the process to step S35. On the other hand, if there are two or more data matching the search condition (“two or more” in step S32), the query execution unit 144 performs context generation processing (FIG. 25) (step S33), and the query execution unit.
- the task generation control unit 152 of 144 executes a task generation process (FIG. 29) (step S34), and advances the process to step S35.
- step S35 the query execution unit 144 executes processing for the page of the DB 206 in the processing step by the task.
- the process for the page of the DB 206 is, for example, a process of reading a page number that matches the search condition if it is an upper page of the index, and a process of reading a Row ID that matches the search condition if it is a leaf page. If it is a page of the table 204, it is a process of reading the column of the record.
- step S36 the query execution unit 144 determines the next page of the DB 206 and the process for the page (step S36), and advances the process to step S37.
- step S37 the query execution unit 144 releases the acquired page of the DB 206 because the processing is completed.
- step S38 the query execution unit 144 determines whether there is a next process. Specifically, the query execution unit 144 determines that the process is “none” when the currently performed process step 73b is completed and there is no next process step in the process block including the process step.
- step S38 If the result of this determination is that there is a next process ("Yes” in step S38), the query execution unit 144 returns the process to step S30, while there is no next process ("No” in step S38). The processing result is passed to the query execution unit 144 (step S39), and the processing step execution process is terminated.
- the query execution unit 144 determines the index root page (the page with the page number “P1” shown in FIG. 6) as the next page of the DB 206, and On the other hand, the search process in the upper page for searching for the key “130” is determined as the process for the page of the DB 206 and the process is started.
- step S30 the query execution unit 144 reads the page P1, and in step S31, searches the page P1 for an entry including “130” in the column c1 (2052).
- the query execution unit 144 acquires one entry (Pt13) including “200” in the column c1 (2052). Therefore, in step S35 and step S36, the next page is the upper page with respect to the page P3.
- the search process is determined as a process for the page of the DB 206.
- the processing for the lower page P3 of the index is performed from step S30 to step S35.
- the query execution unit 144 reads the page P3 from the DB 206, searches the page P3 for an entry including “130” in the column c1 (2052), and enters the page P7 in the entry including “130” in the column c1 (2052).
- the pointer Pt is acquired.
- the query execution unit 144 determines that the page P7 is the next processing target page of the DB 206, and determines that the search processing in the leaf page for the page P7 is processing for the page of the DB 206.
- the query execution unit 144 reads the page P7 from step S30 to step S33, and acquires an entry (E72) including “130” in the column c1 (2052) on the page P7 as shown in FIG.
- the query execution unit 144 performs context generation processing in order to process 9 data other than the data to be processed by the task. (Step S33) is performed, and task generation processing (Step S34) is performed.
- the data processed in the task is the first data
- step S36 the page P21 of the Part table 2051 is determined as the target page of the next DB 206, and as shown in FIG.
- the process for acquiring the record in slot number 2 is determined as the process for the page of DB 206.
- the DBMS 141 With the above processing, the DBMS 141 generates a context 1530 for processing data matching the search condition from the page of the first DB 206 to be processed as shown in FIG. 18, and generates a task from this context 1530. Can be executed as multiple tasks.
- FIG. 25 is a flowchart of the context generation process according to the first embodiment. This process is a process performed in step S33 of FIG.
- the query execution unit 144 performs tasks generated from the initial context 1530 illustrated in FIG. 18 on storage areas # 1 to # 4 that issue I / O requests, an I / O pattern, an I / O size, and the like. And classifying by CPU cost (step S41).
- a task generated from the context 1530 created by the processing of the leaf page (P4 ⁇ ) is generated based on the RowID stored in the unprocessed data list 1562 of the context 1530.
- the query execution unit 144 classifies the RowID held in the context according to the storage area name that issues the I / O request, the I / O size that is the size of the I / O request, the I / O pattern, and the CPU cost. For each unprocessed data list 1562, contexts 1530-1 to 1530-n are generated.
- the query execution unit 144 issues an I / O request to the storage area # 1 with reference to the DB area management table 147 in FIG. To do.
- the query execution unit 144 is processed from the leaf page, so the I / O size 1536 is the page size of the DB 206 (4KB is assumed here) And the I / O pattern 1537 is random.
- the query execution unit 144 acquires the CPU cost with reference to the cost table 1431 of FIG. Since the query execution unit 144 may perform a task generated from the initial context 1530 from processing step # 1 to processing step # 6, the CPU cost 1538 of the initial context 1530 is the CPU cost 1433 of the cost table 1431. The total is 125.
- the query execution unit 144 also executes the above processing for the remaining RowID 20411 shown in FIG. 7, and sets the task generated from the storage area #, I / O size, I / O pattern, and CPU cost to the type of resource to be used. Classify accordingly.
- the query execution unit 144 generates contexts 1530-1 to 1530-n for each of the above-described classifications (step 42) and registers them in the context management unit 153.
- the first context 1530-1 (FIG. 19) accessing the storage area # 1 and the second context 1530-1 (FIG. 20) accessing the storage area # 2. Then, a third context (not shown) for accessing the storage area # 3 and a fourth context (not shown) for accessing the storage area # 4 are generated.
- FIG. 26 is a flowchart of DB page acquisition processing according to the first embodiment. This process is a flowchart showing an example of the process performed in step S30 of FIG.
- the DB buffer management unit 146 searches for a buffer page (a page held in the DB buffer 1460) corresponding to the page of the DB 206 to be acquired (step S51), and whether there is a DB buffer page corresponding to the page of the DB 206 to be acquired. Is determined (step S52).
- step S53 When reading from the external storage apparatus 200 is completed (“completed” in step S53), the DB buffer management unit 146 ends the DB page acquisition process. On the other hand, if reading from the DB 206 has not been completed (“incomplete” in step S53), the process proceeds to step S56.
- step 52 determines whether there is no DB buffer page corresponding to the page of the DB 206 to be acquired (“NO” in step S52)
- the DB buffer management unit 146 selects an empty DB buffer page from the DB buffer 1460. Obtain (step S54). Then, the DB buffer management unit 146 issues a page read request to the DB 206 in order to read a page to be acquired into an empty DB buffer page (step S55), and the process proceeds to step S56. As a result, the acquisition target page is read from the DB 206 into the empty DB buffer page acquired from the DB buffer 1460.
- step S56 the DB buffer management unit 146 waits for page reading from the DB 206 to be completed.
- the DB buffer management unit 146 employs either a synchronous I / O that waits until the reading of the page is completed or an asynchronous I / O that does not complete the reading of the page and executes other processing. be able to.
- the DB buffer management unit 146 interrupts the processing of the task being executed to enter a wait state, and moves the task execution state information 73 to the wait list 1452. Then, the DB buffer management unit 146 determines completion of reading of the acquisition target page by another task. When the DB buffer management unit 146 determines that the reading of the page is completed in the other task, the DB buffer management unit 146 moves the task execution state information 73 of the task to the executable list 1451 and resumes the processing of the task. May be.
- the DB buffer management unit 146 can execute other tasks without waiting for the completion of reading of the page, thereby improving the processing capability of the DBMS 141. Can do.
- the DB buffer management unit 146 ends the DB page acquisition process.
- FIG. 27 is a diagram illustrating an example of a system performance threshold table according to the first embodiment.
- the system performance threshold table 154 holds thresholds for determining whether or not the use of the CPU resource and the I / O resource is sufficient.
- the CPU resource threshold is a CPU usage rate 1541 that is the usage rate of all the processors installed in the computer 100.
- the threshold of the I / O resource includes a disk transfer rate 1542 (unit: MB / s) that is a data transfer amount per unit time from the external storage apparatus 200, and an I / O request process per unit time of the external storage apparatus 200.
- the number is IOPS 1543 (unit: IOPS).
- the packet transfer rate 1544 (unit: pps (Packet Per Second)), which is the number of packets transmitted / received per unit time with other computers, may be added to the I / O resource threshold. it can.
- the packet transfer rate is used as an I / O resource for communication with other computers.
- the network transfer rate (unit: MB / s), which is the data transfer amount with other computers, is set as the packet transfer rate. It may be used instead.
- the thresholds in the system performance threshold table 154 may be specified by the user, or may be automatically calculated by the DBMS 141 from the configuration of the computer system. Alternatively, the threshold value may be obtained by executing a test query for measuring performance, or performing simple CPU processing, simple random READ, or sequential READ.
- the CPU usage rate 1541 is 90% or more of the threshold value, it is determined that the CPU resource is sufficiently utilized.
- the disk transfer rate 1542 is equal to or higher than the threshold value of 2000 MB / s or the IOPS 1543 is equal to or higher than the threshold value of 60000 IOPS, it is determined that the I / O resource is sufficiently utilized.
- the packet transfer rate 1544 is not considered. Therefore, “ ⁇ 1” is set in the packet transfer rate 1544.
- FIG. 28 is a diagram illustrating an example of the performance data table 155 according to the first embodiment.
- the performance data table 155 holds current values for performance data corresponding to the threshold values registered in the system performance threshold value table 154.
- the performance data table 155 includes a CPU utilization rate 1551 that is the utilization rate of all the processors installed in the computer 100 as performance data of the CPU resource.
- the performance data table 155 includes, as performance data of I / O resources, a disk transfer rate 1552 that is a data transfer amount per unit time from the external storage apparatus 200 and an I / O per unit time of the external storage apparatus 200. It includes IOPS1553 which is the request processing number.
- the performance data table 155 also includes a packet transfer rate 1554 that is a data transfer amount with other computers. These values may be calculated each time the DBMS 141 holds the CPU usage time and I / O command history and refers to the performance data table 155. Alternatively, the DBMS 141 may hold the CPU usage time and I / O command history and update the values in the performance data table 155 at regular intervals. In addition, values calculated from values output at regular intervals from an OS command (mpstat command or iostat command) running on the computer 100 may be set in the performance data table 155.
- OS command mpstat command or iostat command
- each value of the performance data table 155 a value acquired by the DBMS 141 or the OS (not shown) of the computer 100 at a predetermined cycle can be used.
- FIG. 29 is a flowchart of task generation processing according to the first embodiment. This process is executed by the task generation control unit 152 of the query execution unit 144 in step S34 of FIG.
- the task generation control unit 152 adjusts the generation of tasks according to the usage status of CPU resources and I / O resources.
- the task generation control unit 152 also adjusts the number of tasks based on the initial task number and the upper limit task number or the lower limit task number set by the query execution unit 144 in step S3 of FIG.
- the number of tasks is the number of tasks existing at that time in the execution task management unit 145.
- the task generation control unit 152 determines whether or not the number of initial tasks is 1 or more (step S59), and if it is 1 or more, generates an initial task (step S67). At this time, the task generation control unit 152 compares the number of tasks with the number of initial tasks (step S68), and sets the initial number of tasks to 0 if the number of tasks is the same as the number of initial tasks (step S69). On the other hand, if the number of tasks is different from the initial number of tasks, the task generation process ends. As a result, the task generation control unit 152 can increase the number of tasks up to the initial number of tasks regardless of the usage status of the CPU resources and I / O resources.
- the task generation control unit 152 compares the lower limit task number with the task number (step S60). If the number of tasks is less than the lower limit task number, the task generation control unit 152 generates an initial task (step S66).
- the task generation control unit 152 determines whether or not an initial task can be generated according to the usage status of CPU resources and I / O resources. Specifically, it is determined whether or not the CPU usage rate 1551 in the performance data table 155 is smaller than the CPU usage rate 1541 in the system performance threshold table 154 (step S61). If the CPU usage rate 1551 in the performance data table 155 is equal to or higher than the threshold CPU usage rate 1541, the task generation control unit 152 ends the task generation processing.
- the task generation control unit 152 determines whether or not the disk transfer rate 1552 in the performance data table 155 is smaller than the disk (or data) transfer rate 1542 in the system performance threshold table 154 (step S62). If the disk transfer speed 1552 of the performance data table 155 is equal to or higher than the threshold disk transfer speed 1542, the task generation control unit 152 ends the task generation process.
- the task generation control unit 152 determines whether or not the IOPS 1553 in the performance data table 155 is smaller than the IOPS 1543 in the system performance threshold table 154 (step S63). If the IOPS 1553 in the performance data table 155 is equal to or greater than the threshold IOPS 1543, the task generation control unit 152 ends the task generation process.
- the task generation control unit 152 determines whether or not the packet transfer rate 1554 in the performance data table 155 is smaller than the packet transfer rate 1544 in the system performance threshold table 154 (step S64). If the packet transfer rate 1554 in the performance data table 155 is equal to or higher than the threshold packet transfer rate 1544, the task generation control unit 152 ends the task generation process.
- the task generation control unit 152 compares the upper limit task number with the current task number (step S65), and when the task number is equal to or greater than the upper limit task number, the task generation process ends. On the other hand, when the number of tasks is smaller than the upper limit number of tasks, the task generation control unit 152 generates an initial task (step S66).
- both the use of the CPU resource and the use of the I / O resource are checked, but only the check of the use status of the I / O resource may be used. Alternatively, it is only necessary to check the usage status of the CPU resource.
- the I / O resource utilization status may be checked only for IOPS, or only for the disk transfer rate, and only for the packet transfer rate. The threshold value may be compared with a combination of these performance data.
- FIG. 30 is a diagram illustrating an example of the storage area performance data table 157 according to the first embodiment.
- the storage area performance data table 157 holds an index of the I / O resource usage status for each storage area in order to determine whether there is a bias in the usage status of I / O resources between the storage areas # 1 to # 4. .
- the storage area performance data table 157 includes an entry including a storage area name 1571, an index 1572, and a value 1573.
- As the index 1572 for example, an outstanding I / O number 1574, which is the number of I / O requests being issued, a disk transfer rate 1575, and an IOPS 1576 are held. These values are set by the same method as the performance data table 155 shown in FIG.
- FIG. 31 is a flowchart of the context acquisition process according to the first embodiment. This process is a process performed in step S21 of FIG.
- the context management unit 153 of the query execution unit 144 preferentially uses the free storage area. Select a context to use. As a result, it is possible to generate a task that utilizes the free storage areas # 1 to # 4.
- the number of standing I / Os 1574 shown in FIG. 30 is used as an index indicating the usage status of I / O resources.
- the usage status of the I / O resource may be determined using the disk transfer rate 1575 or IOPS 1576 shown in FIG. 30 as an index.
- the context management unit 153 refers to the storage area performance data table 157 and selects the storage area name 1571 having the smallest outstanding I / O number 1574 (step S71).
- the context management unit 153 searches the context 1530 in which the storage area name 1571 selected in step S71 is a storage area for issuing an I / O request in descending order of the processing step number (step S72).
- the context management unit 153 searches from the contexts 1530 managed by the context management unit 153. If the corresponding context 1530 exists, the context acquisition process ends (step S73).
- the context management unit 153 searches the contexts 1530 for issuing I / O requests to other storage areas in order from the largest processing step number (step S74).
- the context management unit 153 of the query execution unit 144 preferentially selects a context that issues an I / O request to the storage area name 1571 that uses less I / O resources.
- the query execution unit 144 can generate a task that issues an I / O request to a storage area that uses less I / O resources.
- the context 1530 shown in FIG. 18 is generated.
- an I / O request is made to RowID (P22, 1), RowID (P23, 4), RowID (P24, 2) and storage area # 1. Create a task.
- a storage area name 1571 with a small outstanding I / O number 1574 is selected from the storage area performance data table 157 shown in FIG. 30 and an I / O is added to the selected storage area name 1571.
- a context 1530 for issuing a request is selected to generate a task. Therefore, the query execution unit 144 can generate tasks that issue I / O requests equally to all the storage areas # 1 to # 4.
- the DBMS 140 that dynamically generates a task can generate a task within the thresholds of the system performance threshold table 154 when the usage status of the CPU resource or the I / O resource is within the range.
- the DBMS 140 when the DBMS 140 generates a task, it is possible to improve the utilization rate of the CPU resource and the I / O resource by generating a task that preferentially uses a resource that is insufficiently used. Thereby, it is possible to prevent the processing from being biased toward a specific resource and to improve the processing capability of the DBMS 141.
- Example 2 will be described. At that time, differences from the first embodiment will be mainly described, and description of common points with the first embodiment will be omitted or simplified.
- FIG. 32 is a diagram illustrating an example of the storage area performance threshold table 154A according to the second embodiment.
- the storage area performance threshold value table 154A holds a performance threshold value for determining that the performance is sufficient for each storage area. These threshold values may be set by a method similar to the threshold value setting in the system performance threshold value table 154 shown in FIG. 27 of the first embodiment.
- the storage area performance threshold table 154A includes an entry including a storage area name 1541A, an item 1542A, and a value 1543A.
- a threshold value of the disk transfer speed 1544A and a threshold value of the IOPS 1545A are held.
- FIG. 33 is a flowchart of the context acquisition process according to the second embodiment. This process is a process performed in step S21 of FIG.
- the context management unit 153 of the query execution unit 144 uses the disk transfer rate (threshold value) 1544A and IOPS 1545A (threshold value) of the storage area performance threshold value table 154A of FIG. 33A as indexes.
- the context management unit 153 calculates a disk transfer usage rate, which is an I / O resource usage rate using the disk transfer rate as an index, and an IOPS usage rate, which is an I / O resource usage rate using IOPS as an index (step).
- the disk transfer utilization rate is the disk transfer rate 1552 in the performance data table 155 shown in FIG. 28 of the first embodiment, and the disk transfer for each storage area name 1541A in the storage area performance threshold table 154A shown in FIG.
- the IOPS utilization rate is a value obtained by dividing IOPS 1553 in the performance data table 155 shown in FIG. 28 of the first embodiment by IOPS 1543 in the system performance threshold table 154.
- the context management unit 153 compares the disk transfer usage rate with the IOPS usage rate (step S82). If the disk transfer usage rate is larger, the disk transfer rate is a determining factor of the I / O performance, so the process proceeds to step S83, and the context management unit 153 calculates the disk transfer usage rate for each of the storage areas # 1 to # 4. (Step S83).
- the context management unit 153 selects a storage area with the smallest disk transfer usage rate (step S84).
- step S82 determines whether the IOPS utilization rate is larger, the IOPS is a factor that determines the I / O performance. Therefore, the context management unit 153 sets the IOPS utilization rate for each of the storage areas # 1 to # 4. The calculation is performed (step S85), and the storage area with the smallest IOPS utilization rate is selected (step S86).
- the context management unit 153 searches the storage area selected in step S84 or step S86 for a context that issues an I / O request in descending order of processing steps (step S87), and if the corresponding context exists, selects the context. (Step S88).
- step S88 the context management unit 153 selects contexts in the descending order of processing steps other than the storage area selected in step S84 or step S86 (step S89).
- the I / O resource usage rate is calculated based on the threshold value set for each of the storage areas # 1 to # 4, and the storage areas # 1 to # 4 having the low I / O resource usage rate are calculated.
- a task for issuing an I / O request can be generated.
- Example 3 will be described. At that time, differences from the first embodiment will be mainly described, and description of common points with the first embodiment will be omitted or simplified.
- FIG. 34 is a diagram illustrating an example of the second query of the DB 206 according to the third embodiment.
- the value of the column c1 is “130” and the value of the column c4 is “z” from the Part table 2051 (columns c1 and c2) and the Lineitem table 2054 (columns c3 and c4).
- the column c2 value and column c3 value it means that the column c1 value and the column c4 value are extracted.
- FIG. 35 is a schematic diagram illustrating an example of a query execution plan according to the third embodiment.
- the query execution plan shown in FIG. 35 is an example of the query execution plan generated by the query execution plan generation unit 143 when the DBMS 141 receives the query shown in FIG.
- the query execution plan corresponding to the query shown in FIG. 34 is a processing step in which the DBMS 141 reads all the pages of the DB 206 in the Lineitem table 2054 and obtains a record that meets the conditions.
- # 1 processing step # 2 for performing an index search using the Part index 2041
- processing step # 3 for obtaining a record from the Part table 2051
- data read from the Lineitem table 2054, and data read from the Part table 2051 are hashed
- Processing step # 4 for combining and processing step # 5 for combining these results with a nested loop are included.
- the data in the Lineitem table 2054 is on the Build side
- the data in the Part table 2051 is on the Probe side.
- FIG. 36 is a diagram illustrating an example of the cost table 1431 in which the CPU cost is set for each processing step according to the third embodiment.
- the CPU cost in the processing step may be acquired by the same method as the CPU cost described in FIG. In the illustrated cost table 1431, the CPU cost is set in the same manner as in FIG. 12 of the first embodiment, and the CPU cost on the above-described Build side and the Probe side is set for processing step # 4.
- Example 3 three types of contexts generated by the query execution plan shown in FIG. 35 are assumed.
- processing step # 1 it is assumed that 64 DB pages are managed as one management area, and 16 DB pages make one I / O request.
- the number of pages to be managed may be another management unit, and the number of pages when making an I / O request may be another value.
- FIG. 37 is a diagram illustrating an example of the fifth context 1530-5 according to the third embodiment.
- the fifth context is a context for generating a task that issues an I / O request in the table scan of processing step # 1. Since the I / O request continuously reads 16 pages (4 KB / page), the I / O size 1536 is 64 KB. As described above, since 64 DB pages form one management area, four tasks can be generated in the context 1530-5. For this reason, “4” is set in the number of possible generations 1533.
- the I / O pattern 1537 is sequential because the DBMS 141 reads the pages of the DB 206 continuously.
- the intermediate result 1532 and the execution state 1534 may be set in the same manner as in the first embodiment, and thus the description thereof is omitted here.
- FIG. 38 is a diagram illustrating an example of the sixth context 1530-6 according to the third embodiment.
- the sixth context 1530-6 is a context for generating a task that does not involve an I / O request in the table scan in process step # 1. These tasks are context 1530-6 generated after the task generated from the fifth context 1530-5 completes the I / O request. As long as this context 1530-6 exists, the page of the DB 206 read by the DBMS 141 is Stored in the memory 140.
- 160 is set as the number of possible generations 1533. Since this task does not involve an I / O request, the storage area name 1535 is set to “none”, the I / O size 1536 is set to “0”, and the I / O pattern 1537 is set to “none”. The CPU cost 1538 is set to “20” as in the fifth context 1530-5.
- FIG. 39 is a diagram illustrating an example of the seventh context 1530-7 according to the third embodiment.
- the seventh context 1530-7 is a context for generating a task for performing a random I / O request generated in the processing step # 2. Since this is the same as the context 1530-1 of the first embodiment, description thereof is omitted.
- FIG. 40 is a flowchart illustrating an example of the context acquisition process according to the third embodiment. This process is a process performed in step S21 of FIG.
- the context management unit 153 of the query execution unit 144 calculates the disk transfer usage rate and the IOPS usage rate in the same manner as in the second embodiment, and sets the higher usage rate as the I / O resource usage rate (step S91). .
- the disk transfer usage rate and the IOPS usage rate may be the same as the method in step S81 of the second embodiment.
- the context management unit 153 calculates a CPU usage rate.
- the CPU usage rate is a value obtained by dividing the CPU usage rate 1551 of the performance data table 155 shown in FIG. 28 of the first embodiment by the CPU usage rate 1541 of the system performance threshold table 154 shown in FIG. .
- the context management unit 153 compares the I / O resource usage rate and the disk transfer usage rate (step S92), and determines that if the I / O resource usage rate is low, the I / O resource usage rate needs to be increased. Search is performed preferentially from the context in which I / O is performed in descending order of the processing step number (step S93). For example, in a situation where the fifth context 1530-5, the sixth context 1530-6, and the seventh context 1530-7 exist, the fifth context 1530-5 is selected.
- the context management unit 153 determines that there is no need to increase the I / O resource utilization rate, and prioritizes contexts that do not perform I / O in descending order of processing step numbers. (Step S94). For example, in the situation where the fifth context 1530-5, the sixth context 1530-6, and the seventh context 1530-7 exist, the sixth context 1530-6 is selected.
- the I / O resource usage rate is low by comparing the CPU usage rate and the I / O resource usage rate.
- other methods may be used. For example, a value from which it can be determined that the I / O resource utilization rate is high from the outside may be set.
- a task with a high I / O resource usage rate is generated by selecting a context with a high I / O resource usage rate. Can do.
- I / O resource usage is high, a task with low I / O resource usage rate can be generated by selecting a context with low I / O resource usage rate.
- Example 4 will be described. At that time, differences from the third embodiment will be mainly described, and description of common points with the third embodiment will be omitted or simplified.
- FIG. 41 is a flowchart illustrating an example of context acquisition processing according to the fourth embodiment. This process is a process performed in step S21 of FIG.
- the context management unit 153 of the query execution unit 144 calculates the disk transfer usage rate and the IOPS usage rate in the same manner as in the second embodiment, and sets the higher usage rate as the I / O resource usage rate (step S101). .
- the disk transfer usage rate and the IOPS usage rate may be the same as the method in step S81. Further, the context management unit 153 calculates the CPU usage rate in the same manner as in the third embodiment.
- the context management unit 153 compares the I / O resource usage rate and the CPU usage rate (step S102). If the CPU usage rate is high, the context management unit 153 determines that there is no need to increase the CPU usage rate. A context with a low cost is preferentially selected (step S103). For example, in the situation where the fifth context 1530-5, the sixth context 1530-6, and the seventh context 1530-7 exist, the fifth context 1530-5 or the sixth context 1530-6 is selected. It is.
- the context management unit 153 determines that it is necessary to increase the CPU usage rate, and preferentially selects from the context with the highest CPU cost in descending order of the processing step number (step S104). For example, in a situation where the fifth context 1530-5, the sixth context 1530-6, and the seventh context 1530-7 exist, the seventh context 1530-7 is selected.
- the CPU resource utilization rate (CPU utilization rate) is low by comparing the CPU utilization rate and the I / O resource utilization rate, but it may be determined by other methods. . For example, a value that can determine that the CPU utilization rate is high from the outside may be set. Alternatively, the CPU utilization rate may be compared with a preset threshold value of the CPU utilization rate.
- the DBMS 141 when the usage rate of the CPU resource is low, a task that uses the CPU resource more can be generated by selecting a context with a high CPU cost.
- the DBMS 141 generates a task of the processing step 1432 having the highest CPU cost 1433 with reference to the cost table 1431 when the CPU usage rate is low, and refers to the cost table 1431 when the CPU cost 1433 is high. You may make it produce
- a context that uses a free resource is selected from either the I / O resource or the CPU resource, but it may be compared with a memory resource. Specifically, the I / O resource usage rate and the memory resource usage rate are compared to determine whether or not the memory resource is free. If the memory resource is free, a context that uses more memory resource is determined. You may choose. In addition, the CPU resource utilization rate and the memory resource utilization rate may be compared to determine whether the memory resource is free. If the memory resource is free, a context that uses more memory resource may be selected. .
- Example 5 will be described. At that time, differences from the third embodiment will be mainly described, and description of common points with the third embodiment will be omitted or simplified.
- FIG. 42 is a flowchart illustrating an example of the context acquisition process according to the fifth embodiment. This process is a process performed in step S21 of FIG.
- the context management unit 153 of the query execution unit 144 calculates the disk transfer usage rate and the IOPS usage rate in the same manner as in the second embodiment (step S111).
- the disk transfer usage rate and the IOPS usage rate may be the same as the method in step S81.
- the context management unit 153 compares the disk transfer usage rate and the IOPS usage rate (step S112). If the disk transfer usage rate is high, the context management unit 153 searches for contexts with sequential I / O patterns in descending order of processing step numbers (Ste S113). For example, in a situation where there is a fifth context 1530-5 and a seventh context 1530-7, the fifth context 1530-5 is selected. The context management unit 153 ends if the context can be acquired (step S114).
- the context management unit 153 searches for contexts in which the I / O pattern 1537 is random in descending order of processing step numbers (step S115). For example, in a situation where there is a fifth context 1530-5 and a seventh context 1530-7, the seventh context 1530-7 is selected.
- the context management unit 153 compares the disk transfer usage rate and the IOPS usage rate (step S112). If the IOPS usage rate is high, the I / O pattern 1537 searches for a random context in descending order of processing step numbers. (Step S116). If the context management unit 153 acquires the context, the process ends (step S117).
- the context management unit 153 searches for the contexts in which the I / O pattern 1537 is sequential in descending order of the processing step number (step S118).
- the current I / O pattern is determined by comparing the disk transfer usage rate and the IOPS usage rate. That is, it is sequential when the disk transfer utilization rate is high, and is random when the IOPS utilization rate is high.
- Other methods may be used to determine the I / O pattern.
- the DBMS may store a certain number of I / O pattern histories and determine the I / O pattern.
- the OS may determine the I / O pattern, and the DBMS 141 may refer to the result.
- a context for performing the same I / O pattern can be selected according to the I / O pattern, and a task for performing the same I / O pattern can be generated.
- Example 6 will be described. At that time, differences from the first embodiment will be mainly described, and description of common points with the first embodiment will be omitted or simplified.
- FIG. 43 is a block diagram illustrating a configuration of a computer system according to the sixth embodiment.
- Embodiment 6 shows an example of a database management system in which DBMSs 141-1 to 141-4 operating on a plurality of computers 100-1 to 100-4 perform processing in a coordinated manner.
- FIG. 43 shows an example in which a database management system is configured by four computers (computer 1, computer 2, computer 3, computer 4) 100-1 to 100-4.
- the four computers 100-1 to 100-4 are connected by a network 300.
- Each of the computers 100-1 to 100-4 has an external storage device 200-1 to 200-4, and the computers 100-1 to 100-4 and the external storage device 200-1 to 200-4 are connected by a network.
- the external storage apparatus 200-1 connected to the computer 1 (100-1) includes the storage area # 1
- the external storage apparatus 200-2 connected to the computer 2 (100-2) includes the storage area # 2.
- the external storage device 200-3 connected to the computer 3 (100-3) includes the storage area # 3
- the external storage device 200-3 connected to the computer 4 (100-4) includes the storage area # 3. 4 is included.
- FIG. 43 shows an example in which independent external storage devices 200-1 to 200-4 are configured, but a plurality of computers 100-1 to 100-4 may share one external storage device 200.
- independent storage areas # 1 to # 4 may be provided in the shared external storage apparatus 200.
- the computers 100-1 to 100-4, the DBMSs 141-1 to 141-4, and the external storage apparatuses 200-1 to 200-4 have the same configurations as the computer 100, the DBMS 141, and the external storage apparatus 200 of the first embodiment, respectively. It is.
- FIG. 44 is a diagram illustrating an example of the DB area management table 147A according to the sixth embodiment.
- the DB area management table 147A of the sixth embodiment is different from the DB area management table 147 of the first embodiment in that a DBMS identifier 1474 is added to the DB area management table 147 shown in FIG.
- the other configuration is the same.
- the DBMSs 140-1 to 140-4 in which a plurality of computers 100-1 to 100-4 perform processing in cooperation, the DBMSs 140-1 to 140-4 to be accessed are determined for each of the storage areas # 1 to # 4. Therefore, the DB area management table 147A is obtained by adding a DBMS identifier 1474 to a DB object 1471, a page number 1472, and a storage area name 1473.
- the DBMS identifier 1474 is information for identifying the DBMSs 140-1 to 140-4. With this identifier, it is possible to specify which computer 100-1 to 100-4 is operating.
- FIG. 45 is a diagram illustrating an example of the eighth context 1530-8 according to the sixth embodiment.
- DBMSs 140-1 to 140-4 capable of processing tasks are provided. Since it is limited, a DBMS identifier 1563 for setting a DBMS for executing processing is added to the context 1530-8.
- FIG. 45 shows an example in which the eighth context 1530-8 of the first embodiment is processed by a database management system in which a plurality of computers 100-1 to 100-4 perform processing in a coordinated manner. Since a task that issues an I / O request to the storage area # 1 can be processed only by the DBMS 1 of the computer 100-1, “DBMS 1” is set in the DBMS identifier 1563.
- FIG. 46 is a diagram illustrating an example of the ninth context 1530-9 according to the sixth embodiment.
- FIG. 46 shows an example in which the ninth context 1530-9 shown in the first embodiment is processed by a database management system in which a plurality of computers 100-1 to 100-4 perform processing in a coordinated manner.
- DBMSs 140-1 to 140-4 capable of processing tasks are provided. Limited.
- a DBMS identifier 1563 for setting DBMSs 140-1 to 140-4 for executing processing is added to the ninth context 1530-9.
- the task that issues an I / O request to the storage area # 2 can be processed only by the DBMS 2 executed by the computer 100-2. Therefore, “DBMS2” is set in the DBMS identifier 1563.
- FIG. 47 is a flowchart illustrating an example of task execution processing according to the sixth embodiment. This process is a process performed for a new task in step S17 of FIG. 22 of the first embodiment, and is applied to a database management system in which a plurality of computers 100-1 to 100-4 perform a process in cooperation.
- the task execution process is applied when the query execution unit 144 executes a new task whose process is not determined. A new task whose processing is not decided is generated and processing is started. In order to determine the processing of the new task, context acquisition processing is performed (step S121). The contents of the context acquisition process will be described with reference to FIG.
- the query execution unit 144 determines the DBMS identifier 1563 of the acquired context and compares it with the identifiers of the currently processed DBMSs 140-1 to 140-4 (step S122). For example, when this task execution process is processed by DBMS 1 (140-1) and the eighth context 1530-8 shown in FIG. 45 is acquired, the DBMS identifier 1563 of the context 1530-8 and the DBMS of the own DBMS are obtained. The identifier is the same. In this case, the determination result of step S122 is “true”, and the process proceeds to step S123. On the other hand, when the ninth context 1530-9 in FIG. 46 is acquired, the DBMS identifier 1563 of the context is different from the DBMS identifier of the own DBMS. In this case, the determination result of step S122 is “false”, and the process proceeds to step S125.
- the query execution unit 144 sets task execution state information (step S123). In this process, as in step S22 of FIG. 23, the query execution unit 144 sets task execution state information 73 (see FIG. 13) using the acquired context.
- step S124 The query execution part 144 performs a process step execution process (step S124). This process is the same as step S22 of FIG. 23 shown in the first embodiment, and the query execution unit 144 executes the process step execution process according to the state set in step S123.
- the context acquired to execute the task is transmitted to the DBMS 140-1 to 140-4 of the DBMS identifier 1563 of the context via the network 300 (step S125).
- This context is registered again in the context management unit 153 on the transmitted DBMSs 140-1 to 140-4, a task is generated from the context on the transmitted DBMSs 140-1 to 140-4, and the task is executed.
- FIG. 48 is a flowchart illustrating an example of the context acquisition process according to the sixth embodiment.
- the packet transfer rate 1544 in the system performance threshold table 154 (see FIG. 27) and the packet transfer in the performance data table 155 (see FIG. 28) A value is set for rate 1554.
- the query execution unit 144 compares the packet transfer rate 1544 of the system performance threshold value table 154 with the packet transfer rate 1554 of the performance data table 155 (step S131), and if the packet transfer rate 1554 of the performance data table 155 is small, processing is performed. In order of decreasing step number, the context of the DBMS identifier 1563 different from the own DBMS identifier is preferentially selected (step S132).
- the packet data is preferentially selected from the context of the DBMS identifier 1563 that is the same as the own DBMS identifier in descending order of the processing step number (step S133).
- a task is selected based on one index.
- CPU resources, I / O resources, and memory resources can be used by selecting a task by mixing not only one index but also a plurality of indices.
- a new task can be generated when the memory usage rate is smaller than a specified threshold value.
- the task is generated according to the memory usage rate. When the memory usage rate is smaller than the specified memory usage rate, a task with a large memory resource usage is generated. When the memory usage rate is higher than the specified memory usage rate, a task with a low memory resource usage is generated.
- the context can be identified by the transaction ID, user ID, schema ID, etc., and the order in which tasks are generated may be changed according to the priority of the query.
- the context may be identified by the identifier of the DBMS 141, and the order in which tasks are generated may be changed according to the priority of the DBMS 141.
- the context may be identified by the virtual machine identifier, and the order in which tasks are generated may be changed according to the priority of the virtual machine.
- Example 7 will be described. At that time, differences from the first embodiment will be mainly described, and description of common points with the first embodiment will be omitted or simplified.
- FIG. 49 is a block diagram illustrating a configuration of a computer system according to the seventh embodiment.
- the application server (hereinafter referred to as AP server) 4902 is connected to a computer (hereinafter referred to as DB server) 100 on which the DBMS 141 operates so as to be able to communicate via a communication network 4912.
- the DB server 100 is connected to the external storage device 200 so as to be able to communicate via the communication network 300.
- the user terminal (client terminal) 4901 is connected to the AP server 4902 so that it can communicate via the communication network 4911.
- the DB server 100 executes the DBMS 141 that manages the DB 206 shown in the first embodiment.
- the external storage device 200 stores the DB 206.
- the AP server 4902 executes an AP that issues a query to the DBMS 141 executed by the DB server 100.
- the user terminal 4901 issues a request to the AP executed by the AP server 4902.
- a plurality of user terminals 4901 or AP servers 4902 may exist.
- the AP server management terminal 4903 is connected to the AP server 4902 via the communication network 4914.
- the DB server management terminal 4904 is connected to the DB server 100 via the communication network 4915.
- the storage management terminal 4905 is connected to the external storage device 200 via the communication network 4916.
- the AP server management terminal 4903 is a terminal that manages the AP server 4902.
- the DB server management terminal 4904 is a terminal that manages the DB server 100.
- the storage management terminal 4905 is a terminal that manages the external storage apparatus 200.
- the DB server administrator or user may make settings related to the DBMS 141 from the DB server management terminal 4904. Note that at least two of the management terminals 4903 to 4905 may be common (integrated). Further, at least two of the communication networks 4911, 4912, 4914, 4915, 4916, and 300 may be common (integrated).
- Example 7 for example, the following processing is executed.
- step S141 the user terminal 4901 issues a request (hereinafter referred to as a user request) to the AP server 4902.
- step S142 the AP server 4902 generates a query according to the user request received in step S141. Then, the generated query is issued to the DB server 100.
- step S143 the DB server 100 accepts a query from the AP server 4902 and executes the accepted query.
- the DB server 100 issues a data input / output request (for example, a data read request) necessary for executing the accepted query to the external storage device 200.
- the DB server 100 may issue a plurality of data input / output requests in parallel in the execution of one query. For this reason, the DB server 100 may make the request in step S143 in parallel several times in the execution of one query.
- step S144 the external storage apparatus 200 responds to the DB server 100 with respect to the data input / output request issued in S143.
- the external storage apparatus 200 may perform the response of S144 a plurality of times in parallel.
- step S145 the DB server 100 generates a query execution result and transmits it to the AP server 4902.
- step S146 the AP server 4902 receives the execution result of the query. Then, an answer to the user request received in S141 according to the execution result is transmitted to the user terminal 4901.
- the present invention can be applied even when the AP server 4902 is separated from the DB server 100.
- the configuration of the computer, the processing unit, and the processing unit described in the present invention may be partially or entirely realized by dedicated hardware.
- the various software exemplified in the present embodiment can be stored in various recording media (for example, non-transitory storage media) such as electromagnetic, electronic, and optical, and through a communication network such as the Internet. It can be downloaded to a computer.
- recording media for example, non-transitory storage media
- a communication network such as the Internet. It can be downloaded to a computer.
- the present invention is not limited to the above-described embodiments, and includes various modifications.
- the above-described embodiments have been described in detail for easy understanding of the present invention, and are not necessarily limited to those having all the configurations described.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Operations Research (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
Claims (15)
- ストレージ装置に格納されたデータベースを管理するデータベース管理装置であって、
前記データベースへのクエリを受け付けるクエリ受付部と、
前記受け付けたクエリを実行するために必要な1以上のデータベースオペレーションを表す情報を含むクエリ実行プランを生成するクエリ実行プラン生成部と、
前記生成したクエリ実行プランに基づいて前記受け付けたクエリを実行する際に、データベースオペレーションを実行するためのタスクを動的に生成し、前記動的に生成されたタスクを実行するクエリ実行部と、
を有し、
前記クエリ実行部は、
受け付けたクエリの実行に利用されるリソースの利用状況を取得し、
前記生成されたタスクで実行されるデータベースオペレーションの次のデータベースオペレーションを実行する場合には、前記リソースの利用状況に基づいて新たなタスクを生成し、
当該新たなタスクを前記生成されたタスクと並列して実行する、ことを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況としてI/Oリソースの利用状況を用い、
前記I/Oリソースの利用状況として、前記ストレージ装置からの、若しくは前記ストレージ装置へのデータ転送量、又は、前記ストレージ装置へのI/O要求数を取得し、前記データ転送量、又は、前記I/O要求数に基づいて新たなタスクを生成する、ことを特徴とするデータベース管理装置。 - 請求項2に記載のデータベース管理装置であって、
前記I/O要求数は、前記ストレージ装置に発行される前記ストレージ装置からの入力要求数であることを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況としてCPUリソースの利用状況を用い、
前記CPUリソースの利用状況としてCPU利用率を取得し、前記CPU利用率に基づいて新たなタスクを生成する、ことを特徴とするデータベース管理装置。 - 請求項2に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記データ転送量が予め設定されたデータ転送量閾値よりも小さく、かつ、前記I/O要求数が予め設定されたI/O要求数閾値よりも小さい場合に前記新たなタスクを生成することを特徴とするデータベース管理装置。 - 請求項5に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況として前記I/Oリソースに加えてCPUリソースの利用状況を用い、
前記CPUリソースの利用状況としてCPU利用率を取得し、
前記CPU利用率が予め設定されたCPU利用率閾値よりも小さく、かつ、前記データ転送量が前記データ転送量閾値よりも小さく、かつ、前記I/O要求数が前記I/O要求数閾値よりも小さい場合に前記新たなタスクを生成することを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記新たなタスクを生成する際には、前記リソースの利用状況が所定の閾値未満となる空きリソースを利用するタスクを、分類されたタスクから選択し、生成することを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況としてI/Oリソースの利用状況を用い、
前記I/Oリソースの利用状況として、前記ストレージ装置における前記データベースが分割して配置された記憶領域ごとのI/O要求数を取得し、前記I/O要求数が最小の記憶領域に分割して配置された前記データベースのデータに対するI/O要求を発行するタスクを生成することを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況としてI/Oリソースの利用率と、CPU利用率を取得し、前記I/Oリソースの利用率が前記CPU利用率よりも低い場合には、前記I/Oリソースを利用するタスクを生成し、前記I/Oリソースの利用率が前記CPU利用率よりも高い場合には、前記データベース管理装置のプロセッサを利用するタスクを生成することを特徴とするデータベース管理装置。 - 請求項4記載のデータベース管理装置であって、
前記クエリ実行プラン生成部は、
処理ステップごとのCPUコストを予め設定したコスト情報を保持し、
前記クエリ実行部は、
当該CPU利用率と予め設定した閾値とを比較して、前記CPU利用率が前記閾値未満の場合には、前記コスト情報を参照してCPUコストが大きい処理ステップを実行するためのタスクを生成し、前記CPU利用率が前記閾値以上の場合には、前記コスト情報を参照してCPUコストが小さい処理ステップを実行するためのタスクを生成することを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記クエリ実行部は、
前記リソースの利用状況としてI/Oリソースの利用状況を用い、
前記I/Oリソースの利用状況として、I/Oパターンを識別し、前記新たなタスクは、前記取得したI/Oパターンと同一のI/Oパターンであることを特徴とするデータベース管理装置。 - 請求項1に記載のデータベース管理装置であって、
前記データベース管理装置は、ネットワークを介して他の計算機に接続され、
前記クエリ実行部は、
前記リソースの利用状況としてI/Oリソースの利用状況を用い、
前記I/Oリソースの利用状況として、前記他の計算機と送受信するパケット数である転送パケット数を取得し、
前記転送パケット数が予め設定された転送パケット数閾値よりも小さい場合に前記新たなタスクを生成することを特徴とするデータベース管理装置。 - 計算機がストレージ装置に格納されたデータベースを管理するデータベース管理方法であって、
前記計算機が、前記データベースへのクエリを受け付ける第1のステップと、
前記計算機が、前記受け付けたクエリを実行するために必要な1以上のデータベースオペレーションを表す情報を含むクエリ実行プランを生成する第2のステップと、
前記計算機が、前記生成したクエリ実行プランに基づいて前記受け付けたクエリを実行する際に、データベースオペレーションを実行するためのタスクを動的に生成する第3のステップと、
前記計算機が、前記動的に生成されたタスクを実行する第4のステップと、
を含み、
前記第3のステップは、
受け付けたクエリの実行に利用されるリソースの利用状況を取得し、
前記生成されたタスクで実行されるデータベースオペレーションの次のデータベースオペレーションを実行する場合には、前記リソースの利用状況に基づいて新たなタスクを生成し、
前記第4のステップは、
前記新たなタスクを前記生成されたタスクと並列して実行することを特徴とするデータベース管理方法。 - 請求項13に記載のデータベース管理方法であって、
前記第3のステップは、
前記リソースの利用状況としてI/Oリソースの利用状況を用い、
前記I/Oリソースの利用状況として、前記ストレージ装置からの、若しくは前記ストレージ装置へのデータ転送量と、前記ストレージ装置へのI/O要求数と、を取得するステップと、
前記データ転送量が予め設定されたデータ転送量閾値よりも小さく、かつ、前記I/O要求数が予め設定されたI/O要求数閾値よりも小さい場合に前記新たなタスクを生成するステップと、
を含むことを特徴とするデータベース管理方法。 - 計算機でストレージ装置に格納されたデータベースを管理するプログラムを格納した記憶媒体であって、
前記データベースへのクエリを受け付ける第1のステップと、
前記受け付けたクエリを実行するために必要な1以上のデータベースオペレーションを表す情報を含むクエリ実行プランを生成する第2のステップと、
前記生成したクエリ実行プランに基づいて前記受け付けたクエリを実行する際に、データベースオペレーションを実行するためのタスクを動的に生成する第3のステップと、
前記動的に生成されたタスクを実行する第4のステップと、
を前記計算機に実行させるプログラムを格納した非一時的な計算機読み取り可能な記憶媒体であって、
前記第3のステップは、
受け付けたクエリの実行に利用されるリソースの利用状況を取得して、前記生成されたタスクで実行されるデータベースオペレーションの次のデータベースオペレーションを実行する場合には、リソースの利用状況に基づいて新たなタスクを生成し、
前記第4のステップは、
前記新たなタスクを前記生成されたタスクと並列して実行することを特徴とする記憶媒体。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/905,017 US10515078B2 (en) | 2013-08-30 | 2013-08-30 | Database management apparatus, database management method, and storage medium |
PCT/JP2013/073271 WO2015029208A1 (ja) | 2013-08-30 | 2013-08-30 | データベース管理装置、データベース管理方法及び記憶媒体 |
JP2015533885A JP5950267B2 (ja) | 2013-08-30 | 2013-08-30 | データベース管理装置、データベース管理方法及び記憶媒体 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2013/073271 WO2015029208A1 (ja) | 2013-08-30 | 2013-08-30 | データベース管理装置、データベース管理方法及び記憶媒体 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2015029208A1 true WO2015029208A1 (ja) | 2015-03-05 |
Family
ID=52585824
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2013/073271 WO2015029208A1 (ja) | 2013-08-30 | 2013-08-30 | データベース管理装置、データベース管理方法及び記憶媒体 |
Country Status (3)
Country | Link |
---|---|
US (1) | US10515078B2 (ja) |
JP (1) | JP5950267B2 (ja) |
WO (1) | WO2015029208A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20190031132A (ko) * | 2017-09-15 | 2019-03-25 | 삼성전자주식회사 | 대표 입출력 생성기를 통해 스토리지 디바이스를 테스트하기 위한 방법 및 시스템 |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6245700B2 (ja) * | 2014-04-11 | 2017-12-13 | 国立大学法人 東京大学 | 計算機システム、データの検査方法及び計算機 |
US20160259825A1 (en) * | 2015-03-06 | 2016-09-08 | Dell Products L.P. | Discovery of potential problematic execution plans in a bind-sensitive query statement |
JP6690829B2 (ja) * | 2015-08-28 | 2020-04-28 | 国立大学法人 東京大学 | 計算機システム、省電力化方法及び計算機 |
JP6531587B2 (ja) * | 2015-09-16 | 2019-06-19 | 富士通株式会社 | データベースシステム、データベースアクセス方法、データベースアクセスプログラム、及び、情報処理装置 |
JP7485934B2 (ja) | 2020-06-30 | 2024-05-17 | 富士通株式会社 | 情報処理プログラム、情報処理装置及び情報処理方法 |
US12045235B2 (en) * | 2020-12-14 | 2024-07-23 | International Business Machines Corporation | Access path for database optimizer |
JP2022107451A (ja) * | 2021-01-08 | 2022-07-21 | 富士通株式会社 | 情報処理装置、情報処理システム及び帯域制御方法 |
CN113282571A (zh) * | 2021-05-26 | 2021-08-20 | 北京金山云网络技术有限公司 | 数据转移方法、装置、电子设备和存储介质 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002523839A (ja) * | 1998-08-31 | 2002-07-30 | コンピュータ アソシエイツ シンク,インコーポレイテッド | 急速広範囲dbms分析を行う方法および装置 |
US20050131893A1 (en) * | 2003-12-15 | 2005-06-16 | Sap Aktiengesellschaft | Database early parallelism method and system |
JP2007034414A (ja) * | 2005-07-22 | 2007-02-08 | Masaru Kiregawa | データベース管理システム及び方法 |
JP2010539594A (ja) * | 2007-09-17 | 2010-12-16 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 計算集中的なデータベース・ユーザ定義プログラムを付属の高パフォーマンス並列コンピュータ上で実行するシステムおよび方法 |
JP2012014739A (ja) * | 2011-10-12 | 2012-01-19 | Hitachi Ltd | 計算機システム及びデータベース管理システムプログラム |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090043728A1 (en) * | 2007-08-07 | 2009-02-12 | Barsness Eric L | Query Optimization in a Parallel Computer System to Reduce Network Traffic |
US20090043745A1 (en) * | 2007-08-07 | 2009-02-12 | Eric L Barsness | Query Execution and Optimization with Autonomic Error Recovery from Network Failures in a Parallel Computer System with Multiple Networks |
US20090327216A1 (en) * | 2008-06-30 | 2009-12-31 | Teradata Us, Inc. | Dynamic run-time optimization using automated system regulation for a parallel query optimizer |
US8775413B2 (en) * | 2008-06-30 | 2014-07-08 | Teradata Us, Inc. | Parallel, in-line, query capture database for real-time logging, monitoring and optimizer feedback |
US8447754B2 (en) * | 2010-04-19 | 2013-05-21 | Salesforce.Com, Inc. | Methods and systems for optimizing queries in a multi-tenant store |
US9613092B2 (en) * | 2010-12-31 | 2017-04-04 | Microsoft Technology Licensing, Llc | Allocation of tenants to database services |
US9317552B2 (en) * | 2012-01-30 | 2016-04-19 | Memsql, Inc. | Reusing existing query plans in a database system |
US9280583B2 (en) * | 2012-11-30 | 2016-03-08 | International Business Machines Corporation | Scalable multi-query optimization for SPARQL |
US9639573B2 (en) * | 2013-07-22 | 2017-05-02 | Mastercard International Incorporated | Systems and methods for query queue optimization |
-
2013
- 2013-08-30 JP JP2015533885A patent/JP5950267B2/ja active Active
- 2013-08-30 WO PCT/JP2013/073271 patent/WO2015029208A1/ja active Application Filing
- 2013-08-30 US US14/905,017 patent/US10515078B2/en active Active
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2002523839A (ja) * | 1998-08-31 | 2002-07-30 | コンピュータ アソシエイツ シンク,インコーポレイテッド | 急速広範囲dbms分析を行う方法および装置 |
US20050131893A1 (en) * | 2003-12-15 | 2005-06-16 | Sap Aktiengesellschaft | Database early parallelism method and system |
JP2007034414A (ja) * | 2005-07-22 | 2007-02-08 | Masaru Kiregawa | データベース管理システム及び方法 |
JP2010539594A (ja) * | 2007-09-17 | 2010-12-16 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 計算集中的なデータベース・ユーザ定義プログラムを付属の高パフォーマンス並列コンピュータ上で実行するシステムおよび方法 |
JP2012014739A (ja) * | 2011-10-12 | 2012-01-19 | Hitachi Ltd | 計算機システム及びデータベース管理システムプログラム |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR20190031132A (ko) * | 2017-09-15 | 2019-03-25 | 삼성전자주식회사 | 대표 입출력 생성기를 통해 스토리지 디바이스를 테스트하기 위한 방법 및 시스템 |
KR102316207B1 (ko) | 2017-09-15 | 2021-10-25 | 삼성전자주식회사 | 대표 입출력 생성기를 통해 스토리지 디바이스를 테스트하기 위한 방법 및 시스템 |
Also Published As
Publication number | Publication date |
---|---|
JPWO2015029208A1 (ja) | 2017-03-02 |
JP5950267B2 (ja) | 2016-07-13 |
US20160154848A1 (en) | 2016-06-02 |
US10515078B2 (en) | 2019-12-24 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5950267B2 (ja) | データベース管理装置、データベース管理方法及び記憶媒体 | |
US11151137B2 (en) | Multi-partition operation in combination operations | |
JP4659888B2 (ja) | データベース処理システム、計算機及びデータベース処理方法 | |
US20220237184A1 (en) | Dynamically assigning queries to secondary query processing resources | |
US8856079B1 (en) | Application programming interface for efficient object information gathering and listing | |
US20190392047A1 (en) | Multi-table partitions in a key-value database | |
US11636107B2 (en) | Database management system, computer, and database management method | |
US10942774B1 (en) | Dynamic reassignment of search processes into workload pools in a search and indexing system | |
WO2015118865A1 (ja) | 情報処理装置、情報処理システム及びデータアクセス方法 | |
US10366081B2 (en) | Declarative partitioning for data collection queries | |
CN113474764A (zh) | 共享数据库对象上的流 | |
US11385936B1 (en) | Achieve search and ingest isolation via resource management in a search and indexing system | |
US12007997B2 (en) | Metadata search via N-gram index | |
US11188228B1 (en) | Graphing transaction operations for transaction compliance analysis | |
JP6108418B2 (ja) | データベース管理システム、計算機、データベース管理方法 | |
US9870152B2 (en) | Management system and management method for managing data units constituting schemas of a database | |
Rababa et al. | Optimizations for filter-based join algorithms in MapReduce | |
US11947537B1 (en) | Automatic index management for a non-relational database | |
JP7068210B2 (ja) | データベース管理システム、端末装置及び方法 | |
US11586608B1 (en) | Handling requests to access separately stored items in a non-relational database | |
US11675735B1 (en) | File transfer prioritization during replication | |
US11615092B1 (en) | Lightweight database pipeline scheduler | |
Dory | Study and Comparison of Elastic Cloud Databases: Myth or Reality? |
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: 13892488 Country of ref document: EP Kind code of ref document: A1 |
|
ENP | Entry into the national phase |
Ref document number: 2015533885 Country of ref document: JP Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14905017 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 13892488 Country of ref document: EP Kind code of ref document: A1 |