WO2014077451A1

WO2014077451A1 - Network distributed file system and method using iscsi storage system

Info

Publication number: WO2014077451A1
Application number: PCT/KR2012/011079
Authority: WO
Inventors: 오승; 신중수
Original assignee: 주식회사 유투엔
Priority date: 2012-11-13
Filing date: 2012-12-18
Publication date: 2014-05-22
Also published as: KR101470857B1; KR20140061194A

Abstract

The present invention provides a system and method capable of increasing the efficiency of a distributed file system using an iSCSI storage system in an asymmetric cluster distributed file system. The network distributed file system according to the present invention is driven by including a metadata server (M), a data server (D), and an application server (O), and further includes an extended-type iSCSI storage system including a plurality of iSCSI gateways for processing SCSI commands on the basis of an iSCSI interface using the metadata server (M) and the data server (D) as a block input and output storage space. The iSCSI gateway includes a SCSI target core functioning as a standard interface of a block data input and output that requests from the outside of a network based storage, a SCSI CDB processing module for supporting an iSCSI interface to the distributed file system, and a block layout manager managing data block information distributed on a plurality of physical disk spaces.

Description

【Specification】

[Name of invention]

Network Distributed File System and Method Using iSCSI Storage System

The present invention relates to a system and method for improving the efficiency of a distributed file system using an iSCSI storage system in an asymmetric cluster distributed file system. Background Art

Recently, with the rapid development of the Internet technology, the transmission of multimedia data such as pictures and videos is increasing dramatically. Therefore, a company that provides a multimedia data transmission service using the Internet is generating a very large amount of data. Therefore, it is urgent to introduce a system capable of effectively distributing and managing a large amount of data.

Recently, technological advances in storage systems or file systems have greatly improved the scalability and performance of storage systems. In terms of the file system structure, several systems are making efforts to increase the scalability and performance of distributed storage systems by constructing a so-called asymmetric cluster file system that separates the data I / O path of the file from the file metadata management path. .

This structure allows the client system to directly access the storage devices and can also increase storage scalability by avoiding bottlenecks caused by frequent file access. Enterprise-class storage solutions such as IBM's StorageTank, Panasas 'ActiveScale Storage Cluster, Cluster Filesystems' Luster, Hadoop's DFS, and Google's Google Filesystem have been developed based on this architecture.

In this network-based distributed file system environment, application servers, metadata servers, and data servers communicate over a network to provide input and output of data. In addition, in order to access a specific file, the application server first obtains the location information of the block in which the actual data of the file is stored from the metadata server, and then accesses the data server storing the actual data using the location information. Get the data in the block.

However, in the conventional asymmetric distributed file system, only file 10 is provided and block 10 is not provided.

[Detailed Description of the Invention]

[Technical problem]

An object of the present invention is to provide a distributed file system capable of supporting Block 10 as well as file 10 by solving the problems of the conventional asymmetric cluster distributed file system.

Technical Solution

A network distributed file system running including a metadata server (M), a data server (D), and an application server (0),

The distributed file system blocks the metadata server (M) and data server (D). Expanded iSCSI storage system including a plurality of iSCSI gateways and an iSCSI control server to process SCSKSmall Computers System Interface (SCSKSmall Computers System Interface) commands based on iSCSKinternet Small Computers System Interface (LSC) interface Including

The iSCSI gateway includes a SCSI target core serving as a standard interface for block data input and output requested from outside the network-based storage, a SCSI middleware for supporting an iSCSI interface to a distributed file system, and a plurality of physical disks. And a block layout manager for managing data block information distributed in a space.

In addition, it is preferable that the control server includes a function of configuring an iSCSI volume using a storage space of a distributed file system and controlling a plurality of iSCSI gateways.

In addition, the iSCSI target core may include iSCSI target management, network portal management, target portal group management, LUN mapping, LUN Handshake Authentication Protocol (CHAP) based iSCSI authentication management, multiple connection type iSCSI session management, iSCSI PDU (Protocol Data Unit) and Command processing feature It is characterized by having at least one function.

In addition, the SCSI middleware is a SCSI CDB module for processing SCSI CDB (Command Descriptor Block), decode the CDB to process the command according to the standard of the SCSI command, and in conjunction with the block layout manager, data input Or has a function of commanding to output.

In addition, the block layout manager manages the location, storage unit, and number of file chunks constituting the iSCSI volume, and the location, storage unit, and count information of the file chunks and the block I / O received from the SCSI CDB processing modules. It has a function of performing block data I / O by mapping block address and I / O size information which are parameters.

The metadata of the file managed by the metadata server may include at least one of a file name, a directory structure, a data fragment list constituting the file, and location information on the data server.

In addition, the network distributed file method using the iSCSI storage system is driven by the network distributed file system using the iSCSI storage system,

Requesting, by the server, the iSCSI controller from the metadata server to generate an iSCSI volume;

Assigning, by the metadata server, an ID of an iSCSI volume to be created and assigning chunks to be stored in data servers distributed in a network by a size; And

When the metadata server returns the iSCSI space ID information to the iSCSI management server, the returned iSCSI space ID information may be transmitted to a block layout manager associated with the corresponding iSCSI volume.

In addition, the network distributed file method using the iSCSI storage system is:

From the application server to the iSCSI gateway in the iSCSI standard manner If the SCSI block data is requested to be read, the iSCSI target core analyzes the iSCSI PDU, decomposes the data, and delivers the SCSI CDB to the iSCSI middleware;

Analyzing, by the iSCSI middleware, the SCSI command and transmitting the logical block address (LBA) and read size information of the block to be read to the block layout manager;

The block layout manager calculates the chunk ID and offset configuring the corresponding iSCSI space using the LBA address, delivers the size of the data to be read to the metadata server, and sends the chunk ID and offset to the metadata server. Aligning, in ID order, a list of all the chunk file locations in the data server area distributed in the network based on the size information of the data to be read, to the block layout manager; The block layout manager reads the data as much as the size of data to be read from the offset of the first chunk file to the last chunk file based on the chunk file position list, and writes the data to a buffer located in the block layout manager.

The block layout manager delivers a location in a memory of a buffer to the iSCSI middleware, and the iSCSI middleware generates and transmits a SCSI CDB including the contents of the buffer to the iSCSI target core; And

The iSCSI target core may include generating an iSCSI PDU including a SCSI CDB and transferring the generated iSCSI PDU to an application server in an iSCSI standard manner.

In addition, the network distributed file method using the iSCSI storage system is: The block layout manager writing data to a buffer located in the block layout manager (1-3) by the size of data to be written from the offset of the first chunk file to the last chunk file based on the chunk file position list;

The block layout manager transmits the data write completed to the iSCSI middleware 1-2, and the iSCSI middleware 1-2 generates the SCSI CDB including the data write completed. It may further comprise the step.

Advantageous Effects

By using the distributed file system according to the present invention, since not only the file 10 but also the block 10 can be supported, data transmission and storage can be efficiently performed. [Brief Description of Drawings]

1 is a diagram illustrating a schematic configuration of an iSCSI storage system and a network distributed file system using the storage system according to an exemplary embodiment of the present invention.

FIG. 2 is a diagram schematically illustrating an iSCSI volume formation procedure for supporting block 10. FIG.

FIG. 3 is a diagram schematically illustrating a procedure of processing data read in block 10.

FIG. 4 is a diagram schematically showing a procedure of processing data write in block 10. FIG.

[The best form to carry out invention] An exemplary embodiment of an iSCSI storage system and a network distributed file system using the same according to a preferred embodiment of the present invention will be described below with reference to the accompanying drawings.

The network distributed file system according to the present embodiment is a storage system for generating block iSCSI LUN (LUN) and processing the block 10 by providing it to an iSCSI initiator through an iSCSI target.

In addition, according to a preferred embodiment of the present invention, the expansion of storage capacity is provided in the form of expanding nodes of a distributed file system.

In addition, according to a preferred embodiment of the present invention, the expansion of the overall input / output bandwidth is provided in the form of expanding the iSCSI target of the gateway type.

According to a preferred embodiment of the present invention, the storage availability is secured by using two or more iSCSI gateways, and is provided in a form of fail-over to another iSCSI gateway automatically when one iSCSI gateway fails. .

As shown in FIG. 1, a network distributed file system according to a preferred embodiment of the present invention includes a metadata server (M), a data server (D), and an application for virtualizing a large amount of physically separated storage resources into a single storage space. Contains server (0). Specifically, the metadata server (M) provides control path functionality in a distributed file system. To perform. In addition, the metadata server M manages meta information (file attributes, paths, path lists of data fragments and duplication or not) in the file. In addition, the data server D performs a function of a data path in a distributed file system. In other words, when a file is divided into data fragments by the metadata server M, the functions of distributing and storing the separated data fragments are performed. In addition, the distributed file system according to a preferred embodiment of the present invention is a system for processing SCSI commands based on the iSCSI interface by utilizing the metadata server (M) and the data server (D) as a block input and output storage space. In addition, according to a preferred embodiment of the present invention, the iSCSJ initiator is formed in the application server (0).

In addition, the network distributed file system according to a preferred embodiment of the present invention further includes an scalable iSCSI storage system, which includes a plurality of iSGSI gateways 1 and an iSCSI control server 2. In addition, the plurality of iSCSI gateways 1 includes an iSCSI target core 1-1, a SCSI middleware (1-2), and a block layout manager set (1-3).

In addition, the iSCSI control server (2) and the plurality of iSCSI gateways (1) that make up the scalable iSCSI storage system with the metadata server (M), data server (D) and application server (3) are connected by Ethernet. Connected, and configured to provide a block input / output service to a plurality of existing iSCSI initiators 3 in a scalable manner.

In addition, according to a preferred embodiment of the present invention, the metadata server (M) and the data server (D) used as a base storage of the scalable iSCSI storage system are iSCSI gateways. (1) It is desirable to provide distributed data processing on a per-substrate basis and to provide scalability of block-based storage capacity. In addition, the iSCSI gateway 1 is preferably configured in plurality in order to meet the performance requirements of a plurality of iSCSI initiators using block data input / output services.

On the other hand, the iSCSI gateway 1 according to the preferred embodiment of the present invention is composed of an iSCSI target core (1-1), SCSI middleware (1-2) and block layout manager set (1-3), specifically, It performs the following functions.

First, the iSCSI target core (1 ᅳ 1) functions as a standard interface for block data input and output requested from outside the Ethernet network-based storage. Specifically, the iSCSI target core (1-1) is the iSCSI target management, network portal management, target portal group management, LUN mapping, CHAP (Challenge Handshake Authentication Protocol) based iSCSI authentication management, multiple connection type iSCSI session (Session) ), ISCSI protocol data unit (PDU) and command processing functions. On the other hand, the mapping between iSCSI targets (LUNs) and targets on the UCS FS is achieved through the control of the iSCSI target core of the iSCSI volume manager.

On the other hand, the function of the iSCSI interface is to encapsulate the CDB in the form of encapsulation of SCSI commands and parameters in the form of Ethernet / TCP based on TCP / IP. Only the role of delivering according to the promised procedure is provided. Therefore, SCSI CDB processing modules are essential to support iSCSI interface in distributed file system. SCSI middlewe (1-2) is a module configured for this purpose and decodes the CDB and processes the commands according to the standard of the SCSI command. The SCSI command can be largely classified into a data input / output command and a non-input / output command, and interworking with the block layout manager is determined according to the above classification. In addition, the data input / output command must be linked with the block layout manager. In addition, when the SCSI command processing is completed, the processed result is encapsulated in CDB form and transferred to the iSCSI target core.

The block layout manager 1-3 is a block that manages data block information distributed in a plurality of physical disk spaces. Storage space on the distributed file system is physically divided into data servers and stored as chunks of data. A set of data pieces is mapped to a single file on a distributed file system. Meta information about a file managed by a metadata server of a distributed file system includes a file name, a directory structure, a list of data fragments constituting the file, and location information on the data server. The iSCSI volume (LUN) of an extended iSCSI storage system corresponds to one file designated for storage of a distributed file system. Therefore, in order to connect (LUN mapping) the file and the iSCSI target, a block layout manager that manages meta information of the iSCSI volume connected to the iSCSI target serviced by each iSCSI gateway is essential. The block layout manager manages the location, storage unit, and number of file chunks that make up the iSCSI volume through the distributed file system's metadata server and manages it in internal storage. Block data input by mapping block information and I / O size information which are parameters of block I / O received from SCSI middleware It is responsible for performing the output.

Also, according to a preferred embodiment of the present invention, the iSCSI control server 2 performs control and monitoring of the iSCSI gateway 1. Specifically, the iSCSI server 2 is a block for configuring iSCSI balls and controlling a plurality of iSCSI gateways 1-1 using storage space of a distributed file system. The iSCSI controller server (2) performs iSCSI target management control, network portal management control, target portal group management control, LUN mapping control, and iSCSI authentication management control through the interface provided by the iSCSI gateway (1-1). Can be. ISCSI volume formation procedure to support block IO

First, the iSCSI volume forming procedure for supporting block 10 will be described in detail below with reference to FIG. 2.

First, the iSCSI management server (2) requests the creation of 1 iSCSI volume from the metadata server (M). (ISCSI volume name / size)

Next, the metadata server (M) assigns the ID of the iSCSI ball-flop to be created and assigns chunks to be stored in the data server (D) distributed on the network by the size (Chunk ID, Chunk location, existence). Whether).

Next, when the metadata server (M) returns the iSCSI volume ID information to the iSCSI management server (2), the returned iSCSI volume ID information is passed to the block layout layout manager (1-3) associated with the iSCSI volume. ISCSI volume formation to support block 10 The procedure is complete. Block IO: Data Data read processing

Next, a procedure of processing data read in block 10 is described below with reference to FIG.

First, the application server 0 requests SCSI block data read from the iSCSI gateway 1 in the iSCSI standard manner.

Then, the iSCSI target core (1-1) analyzes the iSCSI PDU, decomposes it, and delivers the SCSI CDB to the iSCSI middleware (1-2).

Subsequently, the iSCSI middleware 1-2 analyzes the SCSI command and transmits the block logical block address (LBA) of the block to read data and the size information of the data to be read to the block layout manager 1-3.

Subsequently, the block layout manager 1-3 calculates the chunk ID and offset constituting the iSCSI name using the LBA address, and transmits the size of the data to be read to the metadata server M.

Subsequently, based on the chunk ID, offset, and size of the data to be read to the metadata server (M), all relevant chunk file position lists of the data server (D) area distributed in the network are sorted in ID order to block the block. Pass it to the layout manager (1–3). Subsequently, the block layout manager (1 to 3) uses the amount of data to read from the offset of the first chunk file to the last chunk file based on the list of chunk file locations. Reads the data and writes the data to the buffer located in the block layout manager (1-3).

The block layout manager 1-3 then transfers the location of the buffer's memory to the iSCSI middleware 1x2, and the iSCSI middleware 1-2 creates a SCSI CDB containing the buffer contents.

Subsequently, the generated SCSI CDB is delivered to the iSCSI target core 1-1, and the iSCSI target core 1-1 generates an iSCSI PDU including the SCSI CDB, and then converts the generated iSCSI PDU into an iSCSI standard method. Pass to application server (0). Block IO: Data Data Write Process

Next, a procedure of processing data write in block 10 will be described below with reference to FIG. 4.

First, the application server 0 requests SCSI block data write to the iSCSI gateway 1 in the iSCSI standard manner.

Subsequently, the iSCSI middleware 1-2 analyzes the SCSI command and transfers the LBA (Logical Block Address) of the block to be written data and the size information of the data write to the block layout manager (1 ᅳ 3).

Subsequently, the block layout manager (1-3) uses the LBA address to determine the corresponding iSCSI volume. Computes the chunk ID and offset constituting the name, and passes the size of the data write to the metadata server (M).

Then, based on the chunk ID, offset, and size of the data write passed to the metadata server (M), the existence of all related chunk files in the data server (D) area distributed in the network is checked. If a chunk is created at the assigned location, the chunk location list is sorted in ID order and then passed to the block layout manager (1-3).

Subsequently, the block layout manager 1-3 writes the data in a buffer located in the block layout manager 1 ᅳ 3 by the size of the data write from the offset of the first chunk file to the last chunk file based on the chunk file position list. .

Subsequently, the block layout manager 1-3 transmits the information on which data writing is completed to the iSCSI middleware 1-2, and the iSCSI middleware 1-2 includes information on which data writing is completed. Create a SCSI CDB.

Subsequently, the generated SCSI CDB is delivered to the iSCSI target core 1-1, and the iSCSI target core 1-1 generates an iSCSI PDU including the SCSI CDB, and then converts the generated iSCSI PDU into the iSCSI standard method. Pass to application server (0). The foregoing has been described in detail with reference to the accompanying drawings an iSCSI storage system and a network distributed file system using the same according to a preferred embodiment of the present invention. However, those skilled in the art to which the present invention pertains to the configuration It will be understood that various modifications and variations are possible. Accordingly, the scope of the invention is limited only by the claims set forth below.

Claims

[Range of request]

[Claim 11

The distributed file system utilizes the metadata server (M) and the data server (D) as a block input / output storage space to process a plurality of iSCSI gateways based on iSCSKintemet Small Computers System Interface (SCSK) small computer systems interface) commands. Further includes an expandable iSCSI storage system including a Gateway and an iSCSI Control Server,

The iSCSI gateway includes a SCSI target core serving as a standard interface for block data input and output requested from outside the network-based storage, a SCSI middleware for supporting an iSCSI interface to a distributed file system, and a plurality of physical disks. A network distributed file system using an iSCSI storage system, which includes a block layout manager for managing data block information distributed in a space.

[Claim 2]

The network distributed file system using the iSCSI storage system of claim 1, wherein the control server comprises a function of configuring an iSCSI volume using a storage space of a distributed file system and controlling a plurality of iSCSI gateways.

[Claim 3]

The method according to claim 1, wherein the iSCSI target core is iSCSI target management, network portal management, target portal group management, LUN mapping, CHAP (Challenge Handshake Authentication Protocol) based iSCSI authentication management, multiple connection type iSCSI session (Session) Network distributed file system using an iSCSI storage system, characterized in that it has at least one function of management, iSCSI protocol data unit (PDU) and command processing.

[Claim 4]

The SCSI middleware of claim 1, wherein the SCSI middleware is a SCSI CDB bundle for processing a SCSI Command Descriptor Block (CDB), decodes the CDB, processes the corresponding commands according to a standard of a corresponding SCSI command, and interworks with a block layout manager. Network distributed file system using an iSCSI storage system, characterized in that it has a function to command to input or output data.

[Claim 5]

The block layout manager of claim 1, wherein the block layout manager manages the location, storage unit, and number of file chunks constituting the iSCSI volume; Network distributed file system using an iSCSI storage system, characterized in that it has a function to perform block data input and output by mapping the block address and the input and output size information parameters of the received block input and output.

[Claim 6]

The method of claim 1, wherein the meta information of the file managed by the metadata server includes at least one of a file name, a directory structure, a data fragment list constituting the file, and location information on the data server. Network distributed file system using storage system.

[Claim 7]

A network distributed file method driven using a network distributed file system using an iSCSI storage system according to any one of the preceding claims:

Assigning, by the metadata server, an ID of an iSCSI volume to be generated and assigning chunks to be stored in data servers distributed in a network by a size; And

And when the metadata server returns the iSCSI volume ID information to the iSCSI management server, the returned iSCSI volume ID information is transmitted to a block layout manager associated with the corresponding iSCSI volume.

[Claim 8]

The method of claim 7, wherein the network distributed file method is:

Upon request to read SCSI block data from the application server to the iSCSI gateway in an iSCSI standard manner, the iSCSI target core analyzes an iSCSI PDU. Disassembling and delivering the SCSI CDB to the iSCSI middleware;

Analyzing, by the iSCSI middleware, a SCSI command and transmitting a logical block address (LBA) of the block to be read data and size information of the data to be read to the block layout manager;

The block layout manager calculates the chunk ID and offset configuring the corresponding iSCSI space using the LBA address, delivers the size of the data to be read to the metadata server, and sends the chunk ID and offset to the metadata server. Aligning, in ID order, a list of all the chunk file locations in the data server area distributed in the network based on the size information of the data to be read, to the block layout manager; The block layout manager reads the data by the size of the data read from the offset of the first chunk file to the last chunk file based on the chunk file position list, and writes the data to a buffer located in the block layout manager; The block layout manager delivers a location in a memory of a buffer to the iSCSI middleware, and the iSCSI middleware generates and transmits a SCSI CDB including the contents of the buffer to the iSCSI target core; And

The iSCSI target core further comprises generating an iSCSI PDU including a SCSI CDB, and delivering the generated iSCSI PDU to an application server in an iSCSI standard manner.

[Claim 9]

8. The method of claim 7, wherein the network distributed file method is: The block layout manager writes data to a buffer located in the block layout manager (1-3) in a size sufficient to write data from the offset of the first chunk file to the last chunk file based on the chunk file position list;

The block layout manager transfers the data write completed to the iSCSI middleware 1-2, and the iSCSI middleware 1-2 creates a SCSI CDB including the data write completed. The network distributed file method further comprises the step of.