Storage device and data processing method
A storage device and memory technology, applied in the direction of electrical digital data processing, special data processing applications, data processing input/output process, etc., can solve the problems of data packet retention, storage device throughput decline, and data packet transmission, etc., to achieve Inhibition of stagnant effect
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
no. 1 Embodiment approach
[0041] figure 1 is a diagram showing an example of the configuration of the storage device according to the first embodiment. The storage device 1 is connected to one or more application servers 2 . The application server 2 is connected to a client terminal (not shown) via a network. The application server 2 can receive a request for the storage device 1 via a network. The application server 2 transmits the request received from the client terminal to the storage device 1 . The request includes, for example, a request for writing data and reading data. Such as figure 1 As shown, the storage device 1 may have a plurality of independent switches 15 and load balancers 16 . There is no need to improve the performance of the switch 15 and the load balancer 16 as the scale of the storage device increases, and it becomes a structure suitable for scale-out.
[0042] The storage device 1 includes a storage unit 10 , a switch 15 , and a load balancer 16 . The storage unit 10 has ...
no. 2 Embodiment approach
[0093] In the second embodiment, a case where a request to write data is input to the storage device 1 is considered. Figure 12 It is a diagram illustrating a case where a plurality of data write requests are input to the storage device 1 . Based on these requests, a plurality of different data packets with write commands are generated by the connection unit 14 . When these requests are write requests for writing data at the same location, each generated data packet arrives at the same destination storage node 11 . In this case, these data packets are not necessarily generated by the same connection unit 14 . In addition, these packets are routed independently. Therefore, these data packets do not necessarily reach the storage node 11 of the transmission destination in the order in which the write requests are input to the storage device 1 . Depending on the order of arrival at the storage node 11 of the transmission destination, data that is requested to be written later ...
no. 3 Embodiment approach
[0107] A disk array device including a plurality of hard disk devices is known. A technique called RAID (Redundant Array of Independent Disks: Redundant Array of Independent Disks) is known as a technique for storing data in a distributed manner in such a disk array device. According to RAID, access processing performance and reliability can be improved by treating a plurality of hard disk devices as one storage device.
[0108] In the third embodiment, RAID (for example, RAID level 5) is applied to the storage device 1 . That is, the storage device 1 generates parity based on a predetermined amount of data, and stores the data and parity in a plurality of storage nodes 11 in a distributed manner. Here, the storage device 1 divides the data (first data) Di (i is a natural number) requested to be written into, for example, two data (second data) Dia and Dib, and generates parity (parity check) based on Dia and Dib. test data) Pi, and save Dia, Dib, and Pi in different storage...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


