Updated on 2023-12-07 GMT+08:00

Storage Pool Types and Performance

DSS provides two types of storage pools, which differ in I/O performance and price. You can select the storage pool type based on your service requirements.

The disk type must be consistent with the storage pool type you selected.

Application Scenarios

  • High I/O storage pool supports only high I/O disks. It can deliver a maximum of 1,500 IOPS per TB and a minimum of 1 to 3 ms read/write latency (single queue, 4 KiB data blocks). This type of storage pools is designed for mainstream high-performance, high-reliability applications, such as enterprise applications, large-scale development and testing, and web server logs.
  • Ultra-high I/O storage pool supports only ultra-high I/O disks. It can deliver a maximum of 8,000 IOPS per TB and a minimum of 1 ms read/write latency (single queue, 4 KiB data blocks). This type of storage pools is perfect for read/write-intensive application scenarios, such as the distributed file systems in the HPC scenarios or NoSQL and relational databases in I/O-intensive scenarios.

Storage Pool Performance

Key metrics of the storage pool performance include read/write I/O latency, IOPS, and throughput.

  • IOPS: Number of read/write operations performed per second
  • Throughput: Amount of data read from and written into a storage pool per second
  • Read/write I/O latency: Minimum interval between two consecutive read/write operations
Table 1 Storage pool performance

Parameter

High I/O

Ultra-high I/O

IOPS

1,500 IOPS/TB

8,000 IOPS/TB

I/O read/write latency (single queue, 4 KiB data blocks)

1 ms to 3 ms

1 ms

Typical application scenarios

Common development and test environments

  • Transcoding services
  • I/O-intensive workloads
    • NoSQL
    • Oracle
    • SQL Server
    • PostgreSQL
  • Latency-sensitive applications
    • Redis
    • Memcache