CN105516365A - Management method of distributed mirror image storage block equipment based on network - Google Patents

Abstract

The invention provides a network-based management method for distributed mirror storage block devices, which relates to the technical field of disk data management, ensures the network intercommunication between the source end and the destination end, creates a resource management group in the same resource management platform, and integrates the source end (local host) and the destination (destination host) are centralized in the same resource management group of the unified resource management platform. In this way, users are allowed to establish a real-time mirror image of local data in units of block devices on a remote machine, so that distributed storage of data blocks can be realized. And add two-way heartbeat detection. When one end pings different peers within 20s, an alarm will be issued and the failure situation at this time will be recorded. It has the advantages of centralized management and intuition.

Description

A management method for network-based distributed mirror storage block devices

technical field

The invention relates to the technical field of disk data management, in particular to a management method of a network-based distributed mirror storage block device.

Background technique

With the large-scale application of disk sharing technology in teaching, experiment and production systems, it plays an increasingly important role in actual production applications, bringing people convenient experience and efficient services. However, with the progress of the times, limited by the single point of failure of shared disks, I/O bottlenecks, etc., the security and scalability of the original shared disks have not met the requirements, so there is an urgent need for safer, more efficient, and better scheme to manage data.

Contents of the invention

In order to solve this problem, the present invention proposes a network-based management method for distributed mirror storage block devices. This method is a storage solution with high reliability, high efficiency and high availability, and has the advantages of centralized management and intuition.

The present invention is based on the network, and the source (local) data is copied to the target (remote) in real time in the form of data blocks through network transmission, without any intervention in the process of copying. In addition, real-time monitoring is carried out on the source and destination ends. When one end fails, an alarm mechanism will be activated to notify the administrator, and the failure time and cause of the failure will be accurately recorded. The administrator can restore it according to the specific fault log, which is very practical and reliable.

Technical scheme of the present invention is:

Through the concept of distributed synchronous mirroring, firstly ensure the network interworking between the source and destination, create a resource management group in the same resource management platform, and concentrate the source (local host) and destination (destination host) on a unified resource management platform in the same resource management group. In this way, users are allowed to establish a real-time mirror image of local data in units of block devices on a remote machine, so that distributed storage of data blocks can be realized. And add two-way heartbeat detection. When one end pings different peers within 20s, an alarm will be issued and the failure situation at this time will be recorded.

At the same time, the data on multiple machines can be copied to the same machine in real time. By creating multiple resource management groups in the management platform, different source machines can be added to different resource management groups, and multiple disks can be divided. The space corresponds to it. In this way, real-time backup of multi-system, multi-place, and multi-platform data is realized, which improves security and achieves high scalability.

The present invention can be used in High Availability (HA), which is similar to a network RAID-1 function. When data is written to the source file system, the data will also be sent to the destination in the network and recorded in a file system in the same form. Data at the source end (primary node) and destination end (standby node) can be synchronized in real time. When the source system fails, the same data will remain on the destination host, which can continue to be used.

The present invention replicates the source (local) data to the target (remote) in real time in the form of data blocks through network transmission, and does not require artificial intervention during the replication process. It also monitors the source and destination in real time. When one end fails, an alarm mechanism will be activated to notify the administrator, and the failure time and cause will be accurately recorded. The administrator can restore it according to the specific fault log, which is very practical and reliable.

The invention provides a more secure and more expandable management method, which can replace the way of using a shared disk array. Because the data exists on the source host and the destination host at the same time, the accident of data loss caused by the failure of the shared disk is avoided. At the same time, data on multiple machines can be replicated to the same machine in real time, achieving high scalability while improving security. When switching, the remote host can continue to serve as long as it uses the backup data of the local end.

Description of drawings

Fig. 1 is a schematic diagram of the management platform of the present invention;

Fig. 2 is a working flow chart of the present invention.

detailed description

The content of the present invention is described in more detail below:

Figure 1 is a schematic diagram of a complete network-based distributed mirror storage block device management platform. The source host disk and the destination host disk perform real-time mirror block synchronization through the network. The entire synchronization is based on block devices. Users can manage The platform monitors the entire synchronization process and handles problems in a timely manner. In this way, functions such as real-time replication of local disks, high security and high availability of remote disaster recovery can be realized.

The management platform mainly uses monitoring and management to ensure distributed mirror storage block devices. The specific implementation methods are as follows:

(1) Format and mount the local disk according to actual needs;

(2) Format the same disk on the destination end as the local end;

(3) Configure the network so that the local end and the destination end can realize network communication;

(4) Install and log in to the management platform, create a new real-time synchronization group, add source and destination resources to the group, and add heartbeat monitoring;

(5) After the configuration is complete, start mirroring. Real-time storage, fault detection, failover and other functions of distributed mirror storage block devices can be realized;

(6) When a problem occurs, the log of the management platform can be used to monitor and deal with the fault in real time.

Claims (3)

1. A management method based on a network-based distributed mirror storage block device, characterized in that, Firstly, ensure the network communication between the source end and the destination end, create a resource management group on the same resource management platform, and gather the source end and the destination end in the same resource management group on the unified resource management platform; in this way, users are allowed to remotely A real-time mirroring of local data in units of block devices is established on the machine, so that distributed storage of data blocks can be realized; and bidirectional heartbeat detection is added. When one end pings different peers within 20s, an alarm will be issued , and record the fault situation at this time; when the source end system fails, the same data will be kept on the destination end host, which can continue to be used. 2. The method of claim 1, wherein, The data on multiple machines can be copied to the same machine in real time. By creating multiple resource management groups in the management platform, different source machines can be added to different resource management groups, and multiple disk spaces can be divided into different resource management groups. Corresponding. 3. The method according to claim 1 or 2, characterized in that, When data is written to the source file system, the data will also be sent to the destination in the network and recorded in a file system in the same form to ensure that the data at the source and destination can be synchronized in real time.