CN112379827A

CN112379827A - Application method for storing double lives

Info

Publication number: CN112379827A
Application number: CN202011139384.7A
Authority: CN
Inventors: 邬玉良
Original assignee: Zhongke Hot Standby Beijing Cloud Computing Technology Co ltd
Current assignee: Zhongke Hot Standby Beijing Cloud Computing Technology Co ltd
Priority date: 2020-10-22
Filing date: 2020-10-22
Publication date: 2021-02-19

Abstract

An application method for storing double lives. The invention relates to the field of computer data processing, realizes the double-active function of data storage, continuously protects data by mirroring copy of the data under the condition of not increasing additional loads of a user side host and storage, and meets the requirement of high availability on shutdown caused by no single-point fault, thereby improving the safety of the data and more effectively protecting the data of a user.

Description

Application method for storing double lives

Technical Field

The invention relates to an application method of a storage double-activity technology, belonging to the field of computer data security processing.

Background

In recent years, people have been in a big data age with the increase of data volume. And the data is more important in the life and work of people.

The data storage is a temporary file generated in the processing process of the data stream or information needing to be searched in the processing process. Data is recorded in a certain format on a storage medium inside or outside the computer. The data store is named, which is to reflect the constituent meaning of the information features. The data flow reflects data flowing in the system and shows the characteristics of dynamic data; the data store reflects data that is static in the system, characterizing static data.

In the storage market, storage classifications (e.g., fig. 1) are classified according to server type: storage of closed systems and storage of open systems. The closed system mainly refers to a mainframe, an AS400 server and the like; an open system refers to a server based on an operating system including Windows, UNIX, Linux, etc. The storage of the open system is divided into: built-in storage and plug-in storage. The plug-in storage is divided into the following steps according to the connection mode: Direct-Attached Storage (DAS) and networked Storage (FAS). The networked storage is further divided into: Network-Attached Storage (NAS) and Storage Area Network (SAN).

In a conventional data center, data of a service system is often stored in a shared storage, for example, a virtualized cluster system and a database dual-computer system share one storage. In the data center in this mode, there is a certain risk of service continuity — if one storage system fails, all service systems connected to the storage system will be halted, and even data will be lost. In order to help customers to deal with the risks, a storage double-live application method is provided, storage systems are mutually mirror images, when one storage system fails and a service is automatically switched to the other storage system, the service can continue to run without being influenced, and data is not lost in the failure process.

Disclosure of Invention

The invention aims to realize the double-active function of data storage, data is processed and stored, so that the data is consistent between two storages, and when one storage is damaged, the data can be accessed through the other storage, thereby improving the safety of the data and more effectively protecting the data of a user.

The storage dual-active mode means that both the main server and the standby server can be read and written for access, and when the source end fails, the standby end can take over the service in time. The traditional storage adopts a centralized type, and data errors are prevented through a two-place three-center storage architecture. The storage architecture of the two-place three-center is a main-standby architecture, and when the main-standby architecture is switched due to faults, the switching speed is slowed down along with the increase of data volume. And the backup does not provide reading and writing to the outside, and whether the backup can be normally used or not can not be ensured after switching. And the storage double-active distributed double-active framework can effectively reduce the influence caused by faults and meet the requirements that RPO is 0 and RTO is less than 1 minute.

When multiple servers access the same data concurrently, the metadata controller MDC prevents two servers from writing to the same disk location at the same time, and ensures that when a certain server reads data content, inconsistent content is not read because other servers are updating the data.

The MDC is responsible for coordinating server access to data, which is located outside of the read and write paths for the data. The client communicates with the MDC through an independent IP link to acquire the position of data and the resource allocation information of the data blocks, and then directly reads and writes the disk in a block-level read-write mode through the SAN network.

Based on the storage dual-activity technology, the method comprises the following steps:

the client application program sends out a write request

Write requests sent to MDC over LAN

The MDC processes the request, responds to which disk blocks the client writes data to through the LAN using live-active techniques

The client end directly writes the data into the storage at the linear speed

Realizing storage double activities

Drawings

FIG. 1 is a block diagram of a storage taxonomy

FIG. 2 is a flow chart for implementing data store liveness

Detailed Description

The overall architecture is as follows:

1. the 4 optical fiber links are used for data synchronization among the arrays, and meanwhile, the heart beat links among the arrays are controlled by A, B to be connected in parallel, and each 2 links are connected.

2. The link between the array and the host is a GE link, and each controller of the array has 2 ports connected with the host to provide front-end services.

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The method comprises the following steps:

1. memory port partitioning

The ports obey the following rules:

1) port usage is largely classified into three categories: a host application port, a liveliness data synchronization port, and an arbitration port. The three types of ports are independent and not shared in the scheme.

2) Parallel networking is a networking mode for copying links recommended in the industry, and the concept of parallel networking is that A, B links between controllers are not subjected to full-interconnection cross processing, and the master-slave end array attribution controllers are consistent.

2. Multi-path configuration

Setting a multipath working module as load balancing:

1) and after the array double-active LUN is mapped to the ESxi host, the array id is inquired.

～esxcli upadm show diskarray

2) Setting load balancing

Esxcli upper graph setup real world model-analysis-p Array _ id is the id queried by show disarray. -p refers to a priority array.

3. Running Internet Explorer at maintenance terminal

4. Enter "https:// XXX.XXX.XXX.XXX:808 in the address bar and Enter the DeviceManager login interface as per the" Enter "system.

5. Creating hard disk domains

The hard disk type of the hard disk domain determines the storage hierarchy that the storage pool can create, and in order to construct the storage pool, the hard disk domain needs to be created first, and the hard disk type and the number of hard disks constituting the hard disk domain need to be specified in the hard disk domain.

6. Creating storage pools

The storage space used by the application server comes from a storage pool of the storage system, so in order to guarantee that the application server can use the storage space of the storage system, the storage pool needs to be created first.

7. Creating LUNs

After a storage pool is successfully created, the storage space of the storage pool cannot be identified by the host. Only after dividing the storage space in the storage pool into a plurality of Logical Units (LUNs) and mapping the LUNs to the host, the host can use the storage space allocated thereto.

8. Creating LUN groups

To facilitate managing and mapping multiple LUNs simultaneously, multiple LUNs may be logically grouped into a LUN group.

9. Creating a host

The scanned hosts can be added to the host list in order to save time when multipath software provided by the manufacturer is installed.

10. Creating a host group

In order to conveniently manage a plurality of hosts at the same time, a plurality of hosts can be logically formed into a host group.

11. Creating double livers

By specifying the dual live domain, the device resources at both ends and the properties of data synchronization, a dual live relationship for the devices at both ends can be created. After the double-activity of the two-end equipment is realized, when a disaster occurs, the arbitration and priority site mechanism provided by the double-activity can judge the sites which continuously provide services for the service, and finally, less data loss and service interruption are realized.

The preconditions are as follows: the two-end equipment can normally communicate and the system works normally, and the local end equipment and the far-end equipment must support the double-active characteristic.

Claims

1. A method for storing double live is characterized in that:

1) a client application program sends a write request;

2) the write request is sent to the MDC through the LAN;

3) the MDC processes the request, responds to the client through the LAN by applying a double-active technology and writes data into the corresponding disk block;

4) the client side directly writes the data into the storage at a linear speed, and storage double-activity is realized.

2. The method of claim 1 wherein the write request is performed simultaneously in a dual disk array with the data times at both ends being precisely coincident.

3. The method of claim 1, wherein the MDC processes the request to synchronize the data written to the disk blocks to maintain a high degree of consistency of the application data.