Archived data storage method, system, device and medium

Abstract

The invention particularly relates to an archived data storage method, a system, a device and a medium. The method comprises the following steps: receiving an archived data write-in request to distribute archived data to a corresponding binary large object, so that the archived data is written into one or more regions of region data regions of one or more disks of the same disk group; and receiving a binary large object sealing request to seal the corresponding binary large object, so that the written region metadata of the one or more regions are written into the region metadata regions of the one or more disks of the same disk group. The method can realize large-capacity storage of archived data, has the characteristics of high reliability and low cost, achieves better balance among storage cost, usability and data retrieval time limit compared with the prior art, and obtains higher data reliability compared with the data in the prior art.

Description

technical field [0001] In particular, the present invention relates to a storage method, system, device and medium for archiving data. Background technique [0002] With the advent of the data age, the demand for massive data storage is increasing. A large proportion of these data belongs to archive data with low access frequency. Raw data itself has this characteristic, such as video surveillance data, and other small files are generally aggregated before being written into the storage system, compressed into a single large file before writing; (2) archived data does not need to be Revise again; (3) archived data generally needs to be stored for a long time, and the frequency of reading after writing is low; (4) the tolerance for the retrieval time limit of archived data is relatively high, but there are still a small number of needs to be retrieved in real time demand. [0003] For storage systems for such archived data, the current mainstream solutions in the industry a...

AI Technical Summary

Problems solved by technology

[0004] The tape library can provide extremely low unit GB storage cost. However, the tape library storage system has the following defects: (1) The initial construction cost is too high. As a complex storage system, the tape library needs to be built once (2) The tape library has high requirements on the space of the computer room, temperature and humidity, etc. Too high or too low temperature and humidity will cause tape adhesion, which will affect data availability and reliability; ( 3) The software and hardware architecture of the tape library is different from that of general servers, and its software development and operation and maintenance work is relatively difficult, and the ease of use is not good; (4) The tape library has great difficulties and challenges in responding to real-time read requests. In a complete tape library storage system, the number of tape drives is limited, so it is impossible to read and write all tapes in parallel, so it is difficult to meet the demand for real-time reading data

The main disadvantage of using a standard object storage system to store archived data lies in its high cost per GB. The disk density of the storage system is restricted; (2) There is a coupling relationship between computing resources such as CPU and hard disk storage resources in the storage system, and it is impossible to flexibly increase the capacity of storage resources without increasing computing resources during system expansion; (3) mainstream The multi-copy technology used by the standard object storage system has a large waste of space. Taking three copies as an example, the disk space usage rate is only 33%; (4) the hard disk used by the mainstream standard object storage system is traditional magnetic recording (CMR, Conventional Magnetic Recording) hard disk, this type of hard disk has a slightly lower storage density than the Shingled Magnetic Recording (SMR, Shingled Magnetic Recording) hard disk, so the unit GB cost is higher than that of the SMR hard disk

[0006] In addition, neither tape libraries nor standard object storage systems can handle silent data corruption well, which poses a high risk of data reliability for archived data that needs to be stored for a long time

[0007] To sum up, there are certain defects in the existing technologies, which cannot fully meet the storage requirements of archived data.

Images