WO2017206754A1

WO2017206754A1 - Storage method and storage device for distributed file system

Info

Publication number: WO2017206754A1
Application number: PCT/CN2017/085338
Authority: WO
Inventors: 郑跃杰
Original assignee: 中兴通讯股份有限公司
Priority date: 2016-05-30
Filing date: 2017-05-22
Publication date: 2017-12-07
Also published as: CN107451138A

Abstract

A storage method and storage device for a distributed file system. The method comprises: aggregating configuration information about all the distributed file systems, the configuration information comprising a policy attribute for storing file data, a mounting point path of the distributed file system and an interface through which the distributed file system communicates with a server; when a write request for the file data is received, searching for the policy attribute matching the file data according to a pre-set storage policy; and calling the interface of the distributed file system corresponding to the policy attribute matching the file data, writing the file data into the corresponding distributed file system through the interface, and recording a mapping relationship between the path of the file data and the mounting point path of the distributed file system.

Description

Distributed file system storage method and storage device

Technical field

The present disclosure relates to the field of communication technologies, for example, to a storage method and storage device of a distributed file system.

Background technique

Network storage devices are used in products such as CRBT and Wireless Application Protocol (WAP) gateways. In order to meet the requirements of high capacity, high throughput and high reliability, the price of network storage devices has also increased exponentially, and the cost of these devices in the entire system often exceeds 50%. Google's Google File System (GFS) is its own file storage technology, and GFS has reduced the cost of the system.

Distributed File System is a general-purpose storage software platform that runs on general-purpose hardware and provides storage platform support for products that need storage services. It provides massive data services (such as multimedia content storage and services) generated by storage, query retrieval and management products. Data storage, etc.). Building a distributed file system on a cheap general-purpose hardware platform has become an inevitable trend in the development of various storage-type businesses. The distributed file system in the distributed file system platform adopts a distributed asymmetric software architecture. The distributed file system implements the separation of functions and features, and is used to complete data access, access and management functions. The distributed file system platform has clear modules and clear interfaces, making it easy to develop or use other products and projects based on this platform.

Distributed file systems are widely used in many fields such as video and mail. However, in the process of using distributed file systems, there are still many complicated scenarios that restrict the use of distributed file systems. For example, in the use of distributed file systems, the high reliability of data is the most concerned aspect of users. A set of distributed file systems will inevitably produce data consistency problems during use. In addition, different distributed file systems have different strategies for storing copies. Different distributed file systems have different focuses on the storage performance of files of different sizes. Some distributed file systems store and read small files per second. The number of operations (Input/Output Operation Per Second, IOPS) is high, and the large file system can be poor; some large file transfers have high bandwidth, but small file storage has low IOPS. Different storage domains between storage domains inside a distributed file system and multiple distributed file systems use different storage namespaces, and file data sharing is difficult. Different sizes of files do not have a unified storage strategy guide Waste of storage space, etc. When users replace existing distributed file systems, they often face a series of technical difficulties such as network replacement, server communication replacement, software replacement, and operation and maintenance replacement.

Summary of the invention

A storage method and storage device for a distributed file system can uniformly manage a distributed file system, fully utilize the storage performance of the distributed file system, improve storage utilization, and improve user experience.

A storage method for a distributed file system, applied to a server, comprising:

Aggregating configuration information of all distributed file systems, the configuration information including a policy attribute of storing file data, a mount point path of the distributed file system, and an interface of the distributed file system to communicate with the server;

When receiving a write request of the file data, searching for a policy attribute matching the file data according to a preset storage policy;

Calling an interface of the distributed file system corresponding to the policy attribute matched by the file data, writing the file data to the corresponding distributed file system through the interface, and recording a path of the file data and a distributed file The mapping relationship of the mount point path of the system.

Optionally, the aggregating configuration information of all distributed file systems includes:

When a distributed file system is detected, the policy attributes of the distributed file system, the mount point path of the distributed file system, and the interface of the distributed file system are included, the distributed file system including the existing distributed file System and new distributed file system;

After the configuration of the distributed file system is completed, the configuration information of all distributed systems is aggregated.

Optionally, the method further includes:

And when the read request of the file data is received, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, and The interface reads the file data in the corresponding distributed file system.

Optionally, the method further includes: when receiving the write request of the file data, encrypting the file data that needs to be written;

The method also includes decrypting the encrypted file data that needs to be read when a read request for the file data is received.

Optionally, the method further includes:

Determining, according to configuration information of all distributed file systems of the aggregation, a distributed file system other than the file data written to the corresponding distributed file system;

The interface of the other distributed file system is invoked to remotely back up file data written to the corresponding distributed file system to the other distributed file system.

Optionally, the method further includes:

When receiving the deletion request of the file data, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, from the corresponding The file data is deleted from the distributed file system;

If the file data has a remote backup in the other distributed system, the file data is deleted from the other distributed file system of the remote backup.

A storage device for a distributed file system, deployed on a server, comprising:

a system extensibility module, configured to aggregate configuration information of all distributed file systems, where the configuration information includes a policy attribute for storing file data, a mount point path of the distributed file system, and an interface for the distributed file system to communicate with the server;

a policy selector configured to: when receiving a write request for file data, find a policy attribute that matches the file data according to a preset storage policy;

a data distribution module, configured to invoke an interface of a distributed file system corresponding to a policy attribute matched by the file data, by using the interface to write the file data into the corresponding distributed file system, and recording file data The mapping between the path and the mount point path of the distributed file system.

Optionally, the system extensibility module is set to:

Optionally, the data distribution module is further configured to:

When receiving the read request of the file data, calling the distributed file system corresponding to the file data according to the mapping relationship between the path of the file data and the mount point path of the distributed file system And reading file data in the corresponding distributed file system through the interface.

Optionally, the device further includes:

a data encryption and decryption module configured to encrypt file data to be written when a write request for file data is received; and to perform encrypted read file data to be read when receiving a read request for file data Decrypt.

Optionally, the data distribution module is further configured to:

A computer readable storage medium storing computer executable instructions arranged to perform the above method.

A server that includes:

At least one processor;

a memory communicatively coupled to the at least one processor; wherein

The memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to cause the at least one processor to perform the method described above.

The distributed file system storage method and the storage device perform unified operation on different distributed file systems by performing aggregation operation on the configuration information of the distributed file system, and perform file data writing, reading, and deleting; pre-configured storage The policy and the writing of the file data according to the storage policy improve the communication storage utilization of the storage server; the file data can be encrypted, so that the file data stored in the distributed file system is invisible to the underlying system, thereby ensuring the file data. safety; You can remotely back up file data to implement remote backup and disaster recovery and ensure the reliability of important file data. Therefore, the distributed file system storage method and storage device fully utilize the distributed file system read and write performance and improve the user experience.

DRAWINGS

The drawings are intended to provide an understanding of the technical solutions, and constitute a part of the specification, and the technical solutions are explained together with the following embodiments, and do not constitute a limitation of the technical solutions.

1 is a schematic structural diagram of a storage device of a distributed file system in an embodiment;

2 is a schematic flowchart of a method for storing a distributed file system in an embodiment;

3 is a schematic diagram showing the steps of a distributed file system storage method in an embodiment;

4 is a schematic diagram showing the hardware structure of a server in an embodiment.

detailed description

In order to make the technical solutions clearer, the embodiments will be described in detail below with reference to the accompanying drawings. The features in the following embodiments and examples may be arbitrarily combined with each other without conflict.

The steps illustrated in the flowchart of the figures may be executed in a computer system such as a set of executing computer executable instructions. Although a logical order is shown in the flowcharts, in some cases, the steps shown or described may be performed in a different order than in the following embodiments.

FIG. 1 is a schematic structural diagram of a storage device of a distributed file system in the embodiment. As shown in Figure 1, the device can be deployed on a server. The file storage is implemented by uniformly scheduling at least two distributed file systems, at least two storage domains, and at least two namespace file systems, so as to achieve high scalability and high availability between at least two distributed file systems. Remote backup, policy generation, and data encryption and decryption.

The storage device of the distributed file system of this embodiment may include: a system extensibility module, a policy selector, and a data distribution module.

The system extensibility module is configured to aggregate configuration information of all distributed file systems, the configuration information including policy attributes for storing file data, a mount point path of the distributed file system, and an interface for the distributed file system to communicate with the server.

The policy attributes of the stored file data may include distribution of large files and small files, cold data and Storage strategy such as distribution of hot data and number of copies. The large file may be a file with a file capacity greater than 1 MB, and the small file may be a file with a file capacity less than or equal to 1 MB; the hot data may be data frequently used by the user, and the cold data may be data with a low frequency of use by the user.

Optionally, the system extensibility module is configured to configure a policy attribute of the distributed file system, a mount point path of the distributed file system, and an interface of the distributed file system when the distributed file system is detected, the distribution File system includes the existing distributed file system and the new distributed file system; and after the configuration of the distributed file system is completed, the configuration information of all distributed systems is aggregated, and the data distribution module is notified for data distribution. The module is uniformly scheduled.

The policy selector is configured to, when receiving a write request of the file data, find a policy attribute that matches the file data according to a preset storage policy.

Optionally, the preset storage policy may be selecting a distributed file system according to the size of the file data or the type of the file data.

For example, for file data of different sizes, small file data is stored in a distributed file system with high IOPS, and large file data is stored in a distributed file system with high bandwidth, thereby fully utilizing the storage performance of the storage system.

For example, for different types of file data, temporary file data can be stored in a single-copy distributed file system, reducing the impact of temporary file data on file system storage performance; important file data can be stored in multiple copies of a distributed file system, or Remote backup and disaster recovery through data distribution modules ensures the reliability of important file data.

The data distribution module is configured to invoke a policy attribute matching the file data corresponding to an interface of the distributed file system, by using the interface to write the file data into the corresponding distributed file system, and recording path information of the file data and The mapping relationship of the mount point path information of the distributed file system, thereby facilitating subsequent file data distribution.

Optionally, the interface that communicates with the data distribution module follows a unified Portable Operating System Interface (POSIX), which defines an interface standard provided by the operating system for the application.

When the client calls the Input Output (IO) interface, the data distribution module can call at least two interfaces of the distributed file system to read and write file data, record the path information of the file data, and mount the distributed file system. Point mapping of path information to facilitate subsequent file data segmentation hair.

According to the needs, for example, according to the importance of the file data, the data distribution module can also back up important file data to other distributed file systems synchronously or asynchronously, thereby realizing remote backup of important file data. The data distribution module may be further configured to: when receiving the deletion request of the file data, deleting the mapping from the corresponding distributed file system according to the mapping relationship between the path of the file data and the mount point path of the distributed file system File data; and if the file data has a remote backup in the other distributed system, the file data is deleted from the remotely backed up distributed file system.

Optionally, the storage device of the distributed file system includes a data encryption and decryption module, and the data encryption and decryption module is configured to encrypt the file data to be written when receiving the write request of the file data; and when the file is received When the data is read, the encrypted file data that needs to be read is decrypted.

Optionally, the data encryption and decryption module provides an encryption and decryption algorithm, and when receiving the write request of the file data, encrypts the file data that needs to be written, and when receiving the read request of the file data, the pair needs to read out The encrypted file data is decrypted, so that the file data stored in the distributed file system is invisible to the underlying system, thereby ensuring the security of the file data.

In this embodiment, the basic configuration of the hardware may be as follows: a central processing unit (CPU) and a solid state drive (SSD). The CPU is an Intel E5-2620, which includes 6 cores, 12 threads, and a CPU clocked at 2500 MHz or higher. The mapping relationship between the path information of the SSD record file data and the mount point path information of the distributed file system, the memory of the SSD is greater than 96 Gbytes (Byte).

Based on the apparatus shown in FIG. 1, as shown in FIG. 2, this embodiment provides a storage method of a distributed file system.

In step 210, configuration information of all distributed file systems is aggregated, the configuration information including policy attributes for storing file data, a mount point path of the distributed file system, and an interface for the distributed file system to communicate with the server.

In step 220, when a write request for file data is received, a policy attribute matching the file data is searched according to a preset storage policy.

In step 230, the interface of the distributed file system corresponding to the policy attribute matched with the received file data is invoked, and the received file data is written into the corresponding distributed file through the interface. In the system, the mapping relationship between the path of the file data and the mount point path of the distributed file system is recorded.

Based on the method shown in FIG. 2, as shown in FIG. 3, this embodiment provides a distributed file system storage method.

In step 310, the system extensibility module aggregates configuration information of all distributed file systems, the configuration information including policy attributes for storing file data, a mount point path of the distributed file system, and an interface for the distributed file system to communicate with the server. .

In step 310, the system extensibility module can configure each distributed file system, including configuring the mount point path, policy attributes, and interfaces for each distributed file system.

The system extensibility module aggregates configuration information of all distributed file systems. When a new distributed file system is found, the configuration information of the newly added distributed file system can be aggregated, and multiple distributed file systems can be configured. There are multiple replica redundancy policies, and the distributed distributed file system can satisfy the POSIX semantic specification.

The data distribution module is notified after the configuration is completed, so that it is uniformly scheduled by the data distribution module.

In step 320, when a write request for file data is received, the file data to be written is encrypted.

In step 320, when the user process receives the write request of the file data, the data encryption and decryption module determines whether the file data needs to be encrypted, for example, the file data is private data, and the data encryption and decryption module can The data is encrypted.

In step 330, a policy attribute matching the file data is searched according to a preset storage policy.

In step 330, the pre-set storage policy may be to select a distributed file system according to file data size or file data type or the like. The policy selector can select a corresponding distributed file system for the file data according to a preset storage policy.

In step 340, the interface of the distributed file system corresponding to the file data is called, and the file data is read and written through the interface.

In step 340, the data distribution module may call an interface of the distributed file system corresponding to the policy attribute matched by the file data, and write the file data into the corresponding distributed file system, for example, write the file data. Into the distributed file system 1.

In step 350, the file data is remotely backed up to other distributed file systems.

In step 350, the data distribution module can asynchronously distribute the file data to other distributed file systems, such as remote backup to the distributed file system 2, to ensure important file data, as needed, for example, the file data is an important file. reliability.

Optionally, determining, according to configuration information of all distributed file systems that are aggregated, that the file data is written to another distributed file system other than the corresponding distributed file system, and calling the interface of the other distributed file system, File data is backed up remotely to other distributed file systems.

In step 360, a mapping relationship between the path of the file data and the mount point path of the distributed file system is recorded.

In step 360, the data distribution module may also record the mapping relationship between the path of the file data and the mount point path of the distributed file system to facilitate subsequent file data distribution.

In step 370, when receiving the read request of the file data, according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, the interface of the distributed file system corresponding to the file data is invoked, and File data is read from the corresponding distributed file system through the interface.

In step 370, when the user process receives the read request of the file data, the data distribution module searches for the mapping relationship between the path of the previously recorded file data and the mount point path of the distributed file system, and determines the corresponding distributed file system. And reading file data from the corresponding distributed file system.

In step 380, the file data that needs to be read is decrypted.

In step 380, if the file data to be read is encrypted data, the data encryption and decryption module decrypts the file data.

In step 390, the file data is read.

In step 390, the file data that needs to be read is returned to the user process.

The distributed file system storage method may further include:

When the user process receives the deletion request of the file data, the data distribution module searches for the mapping relationship between the path of the previously recorded file data and the mount point path of the distributed file system, and determines the corresponding distributed file system, the call and the file. An interface of the distributed file system corresponding to the data, the file data is deleted from the corresponding distributed file system;

If the file data has remote backup to other distributed file systems, the data distribution module needs to Determining, according to configuration information of all distributed file systems of the aggregation, other distributed file systems other than the corresponding distributed file system written by the file data; calling the interfaces of the other distributed file systems, and remotely backing up The file data is deleted in other distributed file systems.

The storage method and the storage device of the distributed file system of the embodiment perform unified operation on at least two distributed file systems by performing aggregation operation on the configuration information of the distributed file system, and perform file data writing and reading. Delete; pre-configure the storage policy and write the file data according to the storage policy, which improves the storage utilization of the server; the file data can be encrypted, so that the file data stored in the distributed file system is invisible to the underlying system, thereby Ensure the security of file data; remotely back up file data to achieve remote backup and disaster recovery, and ensure the reliability of important file data.

Therefore, the storage method and the storage device of the distributed file system of the embodiment fully utilize the read and write performance of the distributed file system, thereby improving the user experience.

The present embodiment provides a computer readable storage medium storing computer executable instructions arranged to perform the method of any of the above embodiments.

This embodiment provides a schematic diagram of a hardware structure of a server. Referring to Figure 4, the server includes:

At least one processor 40, which is exemplified by a processor 40 in FIG. 4; and a memory 41, may further include a communication interface 42 and a bus 43. The processor 40, the memory 41, and the communication interface 42 can complete communication with each other through the bus 43. Communication interface 42 can transmit information. Processor 40 may invoke logic instructions in memory 41 to perform the methods of the above-described embodiments.

In addition, the logic instructions in the memory 41 described above may be implemented in the form of a software functional unit and sold or used as a stand-alone product, and may be stored in a computer readable storage medium.

The memory 41 is a computer readable storage medium and can be used to store a software program, a computer executable program, such as a program instruction or a module corresponding to the method in the above embodiment. The processor 40 executes the functional application and the data processing by executing a software program, an instruction or a module stored in the memory 41, that is, implementing the method in the above method embodiment.

The memory 41 may include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application required for at least one function; the storage data area may store data created according to use of the terminal device, and the like. In addition, the memory 41 may include a high speed random access memory, and may also include Non-volatile memory.

The above embodiments can be implemented by means of software plus a general hardware platform. The above technical solution can be embodied in the form of a software product, which can be stored in a storage medium, such as a read-only memory (ROM), a random access memory (RAM), A disk, an optical disk, etc., comprising one or more instructions for causing a computer device (which may be a personal computer, server communication, or network device, etc.) to perform the methods described in the preceding embodiments or embodiments.

Industrial applicability

The distributed file system storage method and storage device can uniformly manage the distributed file system, improve storage utilization, and improve user experience.

Claims

A storage method for a distributed file system, applied to a server, comprising:

Aggregating configuration information of all distributed file systems, the configuration information including a policy attribute of storing file data, a mount point path of the distributed file system, and an interface of the distributed file system to communicate with the server:

When receiving a write request of the file data, searching for a policy attribute matching the file data according to a preset storage policy;

Calling an interface of the distributed file system corresponding to the policy attribute matched by the file data, writing the file data to the corresponding distributed file system through the interface, and recording a path of the file data and a distributed file The mapping relationship of the mount point path of the system.
The method of claim 1 wherein said aggregating configuration information for all distributed file systems comprises:

When a distributed file system is detected, the policy attributes of the distributed file system, the mount point path of the distributed file system, and the interface of the distributed file system are included, the distributed file system including the existing distributed file System and new distributed file system;

After the configuration of the distributed file system is completed, the configuration information of all distributed systems is aggregated.
The method of claim 1 further comprising:

And when the read request of the file data is received, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, and The interface reads the file data in the corresponding distributed file system.
The method of claim 3, further comprising: encrypting the file data to be written when receiving the write request of the file data;

When a read request for file data is received, the encrypted file data that needs to be read is decrypted.
The method of claim 1 further comprising:

Determining, according to configuration information of all distributed file systems of the aggregation, a distributed file system other than the file data written to the corresponding distributed file system;

The interface of the other distributed file system is invoked to remotely back up file data written to the corresponding distributed file system to the other distributed file system.
The method of claim 5 further comprising:

When receiving the deletion request of the file data, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, from the corresponding The file data is deleted from the distributed file system;

If the file data has a remote backup in the other distributed system, the file data is deleted from the other distributed file system of the remote backup.
A storage device for a distributed file system, deployed on a server, comprising:

a system extensibility module, configured to aggregate configuration information of all distributed file systems, where the configuration information includes a policy attribute for storing file data, a mount point path of the distributed file system, and an interface for the distributed file system to communicate with the server;

a policy selector configured to: when receiving a write request for file data, find a policy attribute that matches the file data according to a preset storage policy;

a data distribution module, configured to invoke an interface of a distributed file system corresponding to a policy attribute matched by the file data, by using the interface to write the file data into the corresponding distributed file system, and recording file data The mapping between the path and the mount point path of the distributed file system.
The apparatus of claim 7, wherein the system extensibility module is configured to:

When a distributed file system is detected, the policy attributes of the distributed file system, the mount point path of the distributed file system, and the interface of the distributed file system are configured, the distributed file system package Including existing distributed file systems and new distributed file systems;

After the configuration of the distributed file system is completed, the configuration information of all distributed systems is aggregated.
The device according to claim 7, wherein the data distribution module is further configured to:

And when the read request of the file data is received, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, and The interface reads the file data in the corresponding distributed file system.
The apparatus of claim 9 further comprising:

a data encryption and decryption module configured to encrypt file data to be written when a write request for file data is received; and to perform encrypted read file data to be read when receiving a read request for file data Decrypt.
The apparatus of claim 7, wherein the data distribution module is further configured to:

Determining, according to configuration information of all distributed file systems of the aggregation, a distributed file system other than the file data written to the corresponding distributed file system;

The interface of the other distributed file system is invoked to remotely back up file data written to the corresponding distributed file system to the other distributed file system.
The apparatus of claim 11 wherein said data distribution module is further configured to:

When receiving the deletion request of the file data, the interface of the distributed file system corresponding to the file data is invoked according to the mapping relationship between the path of the file data and the mount point path of the distributed file system, from the corresponding The file data is deleted from the distributed file system;

If the file data has a remote backup in the other distributed system, the file data is deleted from the other distributed file system of the remote backup.
A computer readable storage medium storing computer executable instructions arranged to perform the method of any of claims 1-6.