KR20140120970A - Method for efficient external sorting in intelligent solid state disk and storage device - Google Patents

Abstract

Disclosed is a method for efficiently executing an external sorting algorithm sorting big data through structural features of an intelligent solid state disk (SSD). The external sorting algorithm using grouping can prevent frequent data access, and can minimize the write operations regarding the big data by managing the data regardless of the number of runs. Accordingly, the present invention can reduce the number of runs needed during an external sorting operation to enable an efficient external sorting operation.

Description

[0001] METHOD FOR EFFICIENT EXTERNAL SORTING IN INTELLIGENT SOLID STATE DISK AND STORAGE DEVICE FOR INTELLIGENT SSD [0002]

The present invention relates to a method for sorting large amounts of data in an intelligent solid state disk (SSD), and more particularly, to a method for efficiently performing an external sorting algorithm using an intelligent solid state disk (SSD).

As technology advances, solid state disks (SSDs) have become more bottlenecks because their internal bandwidth is higher than the host interface bandwidth. These bottlenecks in the host interface have become a major obstacle to processing large amounts of data. In particular, the bottleneck in the host interface is that when performing repetitive data access to a large amount of data, performing a simple operation on the accessed data, a data-dependent application The more obvious it is.

 As a solution to the new bottleneck, an intelligent solid state disk (SSD) has been introduced that gives internal data processing ability to process data in SSD by granting data processing capability to SSD inside.

In order to take advantage of SSD with multi-channel structure, ISSD can process data in parallel by giving data processing capability to each channel.

S. Kim et al., " Fast , Energy Efficient Scan inside Flash Memory SSDs , In Proc . Int'l Workshop on Accelerating Data Management Systems using Modern Processor and Storage Architectures , ADMS, 2011. describes how to handle data scanning internally by mounting additional processors and memory in the SSD.

The present invention provides a method for efficiently performing an external alignment algorithm using features of an intelligent SSD (ISSD) structure.

The present invention provides a method for mass data processing in an ISSD.

The external sorting method for ISSD according to an embodiment of the present invention may include grouping a plurality of runs stored in each channel and performing external sorting.

According to one aspect of the present invention, an external alignment method for an ISSD includes grouping a plurality of runs stored in each channel; Performing data sorting on the groups and transmitting the sorted results to a core memory; And sorting the data received from each channel in the core memory to perform sorting on all data.

According to another aspect, the external alignment method for ISSD further comprises calculating a maximum number of groups that can be generated in one channel, and grouping the plurality of runs stored in each channel into groups, Grouping the plurality of runs stored in each channel based on the maximum number of groups.

According to an embodiment of the present invention, there is provided a data processing system including a run grouping unit for grouping a plurality of runs stored in respective channels, a data sorting unit for performing data sorting for each group, A flash memory controller (FMC) including a transfer unit for transferring the image data; And a core memory for sorting the data transmitted from each channel and performing sorting on the entire data.

로 나타낸다.According to an aspect of the present invention, the maximum number of groups may be calculated using an FMC (Flash Memory Controller) memory of each channel,

Respectively.

로 지정될 수 있다. 상기 2는 상기 코어 메모리에서 상기 각 채널에서 전송받은 데이터를 정렬하여 출력하는 공간을 마련하기 위한 변수이다.According to another aspect, each of the groups is stored in a space of the core memory, and the size per group in the space of the core memory is

. ≪ / RTI > 2 is a variable for providing a space for sorting and outputting data received from each channel in the core memory.

According to another aspect, the core memory can construct a tree according to the number of the channels existing in the intelligent SSD regardless of the number of the runs, and compare the data.

According to an embodiment of the present invention, it is possible to prevent frequent access to data through a method for processing large capacity data in the ISSD, and it is possible to minimize writing operation of large capacity data by managing data regardless of the number of runs, Can be performed by only two write operations, a run generation step and a merging step.

According to an embodiment of the present invention, there is provided a method for efficiently performing external alignment by reducing a number of runs when an external alignment is performed using a structural feature of an ISSD.

1 is a diagram showing the structure of an intelligent SSD (ISSD).
2 is a diagram showing a data transmission path of the intelligent SSD (ISSD).
3 is a diagram illustrating a basic strategy of an external sorting algorithm of an intelligent SSD (ISSD) according to an embodiment of the present invention.
FIG. 4 is a diagram illustrating an external sorting algorithm through grouping in an intelligent SSD (ISSD) environment according to an embodiment of the present invention.
5 is a flowchart illustrating an external sorting method in an intelligent SSD (ISSD) environment according to an embodiment of the present invention.
6 is a block diagram illustrating the functions of a flash memory controller (FMC) according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram showing the structure of an intelligent SSD (ISSD).

Referring to FIG. 1, a flash memory cell 102 has a multi-way structure including a plurality of cells and constitutes one channel 103. One channel 103 includes a flash memory controller 104 for controlling a plurality of flash memory cells 102 and a core 105 for managing a plurality of channels existing in a multi-channel structure. Embodiments of the present invention propose a method for efficiently performing an external alignment algorithm using the features of the intelligent SSD (ISSD) structure as shown in FIG.

2 is a diagram showing a data transmission path of the intelligent SSD (ISSD).

FIG. 2 shows a data transfer path 201 for processing data of the intelligent SSD (ISSD). As shown in FIG. 2, data read from the flash memory cell for data processing is transferred to the flash memory controller 202) or core (203).

 In Intelligent SSD (ISSD), data is not transmitted to the host, so it does not go through the bottleneck point generated from the host interface. If the flash memory controller performs parallel processing, it can reduce the data transfer cost by reading from the flash memory cell to the flash memory controller. When large amounts of data are stored in multiple channels, the amount of data transferred to each channel can be reduced. Therefore, an intelligent SSD (ISSD) can be advantageous for large capacity data processing, and the present invention proposes a method for large capacity data processing in such ISSD.

3 is a diagram illustrating a basic strategy of an external sorting algorithm of an intelligent SSD (ISSD) according to an embodiment of the present invention.

The external sorting algorithm of the intelligent SSD (ISSD) is divided into a subfile creation step and a subfile merging step.

The sub-file generation step uses an internal sorting method, and the sub-file merging step can be divided into a part for managing sub-files in the channel and a part for merging the results processed by the channel in the core.

As shown in FIG. 3, the flash memory controller (FMC) manages a plurality of subfiles stored in a channel in a tree form.

In the core, the channels are managed as a tree, and the records transmitted from each channel are compared using a tree. At this time, the core (CORE) can sort the entire data by selecting and outputting the smallest data in order using the tree, and it is repeatedly performed to perform an external sorting algorithm 301 for the entire data Can be performed.

FIG. 4 is a diagram illustrating an external sorting algorithm 401 through grouping in an intelligent SSD (ISSD) environment according to an embodiment of the present invention.

According to one embodiment of the present invention, tree information 405 management generated by performing grouping 404 of sub-files 404 to efficiently perform an external sorting algorithm 401 through grouping, A group merger (407) method is proposed.

As shown in FIG. 4, since the memory size of the flash memory controller 403 is small in performing the external alignment algorithm in the embodiment of the present invention, there is a limitation in managing all the sub-files stored in the channel, (404), and arranges the tree (405) for each group to perform the sorting operation.

The core 402 stores the tree information, and when the operation is completed up to the last group, the data stored in the root of each tree can be read to request data for each channel by setting the order of the tree for reading the next data.

The core 402 finds and arranges the smallest data among the data transmitted from each channel. Since the data transmitted by each channel in each channel is not guaranteed to be the smallest data in the entire data, the data of the previous groups must be stored until the data transmitted in the last group is confirmed.

In this process, the core 402 merges 407 data from the channels and groups the data of the group. At the same time, the data sorted in the previous group are merged together (407). The data of the previous groups is added to the tree structure 406 created for the channel in order to merge the data for each group 407 so that the height of the tree handled by the core is increased by one.

5 is a flowchart illustrating an external sorting method in an intelligent solid state disk (SSD) environment according to an embodiment of the present invention.

In step 501, a plurality of runs stored in each channel of the flash memory controller may be grouped.

In step 502, groups of the runs stored in each channel are grouped and data sorting is performed for each group.

In step 503, the aligned result is transferred to the core memory.

In step 504, the data transmitted on each channel is sorted to perform alignment on the entire data.

Therefore, in the intelligent solid state disk (SSD) environment, the external sort method using grouping can prevent frequent data access and manage the data irrespective of the number of runs because the number of each channel is sorted using the tree structure only .

6 is a block diagram illustrating the functions of a flash memory controller (FMC) according to an embodiment of the present invention.

The flash memory controller 600 includes a run grouping unit 601 for grouping a plurality of runs stored in each channel, a data arranging unit 602 for performing data sorting on a group basis, (603). The core memory sorts the data transmitted from the channel and performs sorting on the entire data.

로 나타낼 수 있다. To manage the run-by-group in the external sorting algorithm, you need to find the maximum number of groups that can be created in one channel. The maximum number of groups that can be managed by the channel is

.

로 나타내고, 2는 상기 코어 메모리에서 상기 각 채널에서 전송된 데이터를 정렬하여 출력하는 공간을 마련하기 위한 변수이다. 상기 코어 메모리는 상기 런들의 수에 상관없이 존재하는 상기 각 채널의 수에 맞게 트리를 구축하여 데이터를 비교한다.
And since each group must be stored in core memory space, you have to specify the size per group in core space. The size per group in core space is

And 2 is a variable for providing space for sorting and outputting data transmitted from each channel in the core memory. The core memory constructs a tree according to the number of channels existing irrespective of the number of the runs and compares the data.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA) A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave. The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. For example, it is to be understood that the techniques described may be performed in a different order than the described methods, and / or that components of the described systems, structures, devices, circuits, Lt; / RTI > or equivalents, even if it is replaced or replaced.

Therefore, other implementations, other embodiments, and equivalents to the claims are also within the scope of the following claims.

401: External Sorting Algorithm by Grouping
402: Core
403: Flash memory controller
404: Grouping multiple runs
405: tree structure in which a plurality of runs are grouped
406: Tree structure created by channel
407: Data Merger at Core

Claims (13)

1. An external alignment method for an intelligent solid state disk (SSD)
Grouping a plurality of runs stored in each channel to perform external sorting,
External alignment method for intelligent solid state disk (SSD). The method according to claim 1,
The method comprises:
Grouping a plurality of runs stored in each channel into groups;
Performing data sorting on the groups and transmitting the sorted results to a core memory; And
Arranging data received from each channel in the core memory and performing sorting on all data
An external sorting method for an intelligent solid state disk (SSD). 3. The method of claim 2,
The method comprises:
Further comprising calculating a maximum number of groups that can be generated in one channel,
Wherein the step of grouping the plurality of runs stored in each channel comprises:
Grouping a plurality of runs stored in each channel based on the calculated maximum number of groups,
External alignment method for intelligent solid state disk (SSD). The method of claim 3,
Wherein the maximum number of groups is calculated using an FMC (Flash Memory Controller) memory of each channel,
External alignment method for intelligent solid state disk (SSD). 5. The method of claim 4,
The maximum number of groups is

Lt; / RTI >
External alignment method for intelligent solid state disk (SSD). 3. The method of claim 2,
Wherein each of the groups is stored in a space of the core memory and the size per group is designated in a space of the core memory,
External alignment method for intelligent solid state disk (SSD). The method according to claim 6,
The size per group is

Lt; / RTI >
2 is a variable for providing a space for sorting and outputting data received from each channel in the core memory,
External alignment method for intelligent solid state disk (SSD). 3. The method of claim 2,
Wherein the core memory constructs a tree according to the number of the channels existing in the intelligent SSD (Solid State Disk) regardless of the number of the runs,
External alignment method for intelligent solid state disk (SSD). A flash memory controller including a run grouping unit for grouping a plurality of runs stored in each channel into groups, a data arranging unit for performing data sorting for each group, and a transmitting unit for transmitting the sorted results to a core memory FMC: Flash Memory Controller); And
A core memory for sorting data received from each channel and performing sorting on all data,
. ≪ / RTI > 10. The method of claim 9,
The run grouping unit,
Calculates the maximum number of groups that can be generated in one channel,
Grouping a plurality of runs stored in each channel based on the calculated maximum number of groups,
Storage device. 11. The method of claim 10,
The maximum number of the groups is calculated using an FMC (Flash Memory Controller) memory of each channel,

Lt; / RTI >
Storage device. 10. The method of claim 9,
Wherein each of the groups is stored in a space of the core memory,

Lt; / RTI >
2 is a variable for providing a space for sorting and outputting data received from each channel in the core memory,
Storage device. 13. The method of claim 12,
Wherein the core memory constructs a tree according to the number of the channels existing irrespective of the number of the runs and compares the data,
Storage device.

Images