Analyzing the Performance of Back-End Ports and Disks

OceanStor Dorado V6 storage systems support disk and enclosure integration, external SAS disk enclosures, and external smart disk enclosures. This section analyzes the performance of external disk enclosures. Ports of back-end SAS disk enclosures or smart disk enclosures have impact on the performance, which lies in the loop consisting of the disk enclosures. Currently, OceanStor Dorado V6 storage systems support 12 Gbit/s SAS and 100 Gbit/s RDMA ports.

A single port provides limited bandwidth. The bandwidth supported by the ports in a loop must be higher than the total bandwidth of all disks in the disk enclosures that compose the loop. In addition, as the number of disk enclosures in a loop grows, the latency caused by expansion links increases, affecting the back-end I/O latency and IOPS. Based on the preceding two points, if there are sufficient SAS ports:

• Evenly distribute disk enclosures to multiple loops.
• If a single controller has multiple back-end interface modules, distribute the loops to multiple modules instead of using the ports on one module.
• Form a loop using less than two disk enclosures.
• Ensure that disks inserted into each disk enclosure are of the same type and quantity.

You can use OceanStor DeviceManager or CLI to query the disk enclosure IDs and then determine loop connections. The ID of a disk enclosure is in the format of DAEabc (a, b, and c are integers), where a indicates the controller enclosure ID, b indicates the hexadecimal loop ID, and c indicates the ID of an enclosure in the loop. For example, the disk enclosure IDs are DAE000, DAE010, DAE020, DAE021, DAE030, and DAE031. The IDs indicate that there are two 2-enclosure loops (DAE020 and DAE021 compose one loop; DAE030 and DAE031 compose another) and two single-enclosure loops (DAE000; DAE010). The loop connections comply with the preceding performance rules.

To view the performance indicators including the port usage rate, total IOPS, and block bandwidth, use OceanStor DeviceManager or run the CLI command.

Performance indicators of back-end ports vary with different versions. The actual interface prevails.

• Use OceanStor DeviceManager to view the indicators.

  Operation path: Insight > Performance. On the Dashboard page, you can view typical performance indicators of various back-end ports. If you need to query more performance indicators, create a metric chart for the controller as prompted on the Analysis page first. For details about how to create a metric chart, see Creating a Metric Chart.

• Run show performance port on the CLI to view the indicators.

  admin:/>show performance port port_id=CTE0.A0.P1 
  0.Bandwidth(MB/s) / Block Bandwidth(MB/s)     
  1.Throughput(IOPS)(IO/s)                      
  2.Read Bandwidth(MB/s)                        
  3.Read Throughput(IOPS)(IO/s) 
  4.Write Bandwidth(MB/s)                       
  5.Write Throughput(IOPS)(IO/s)                
  6.Average I/O Latency(ms)                     
  7.Max. I/O Latency(ms) 
  8.Max Read I/O Latency(ms)                    
  9.Max Write I/O Latency(ms)                   
  10.Average Read I/O Latency(ms)               
  11.Average Write I/O Latency(ms) 
  12.Average I/O Latency(us)                    
  13.Max. I/O Latency(us)                       
  14.Max Read I/O Latency(us)                   
  15.Max Write I/O Latency(us) 
  16.Average Read I/O Latency(us)               
  17.Average Write I/O Latency(us) 
  Input item(s) number seperated by comma:

Storage media supported by OceanStor Dorado V6 storage systems are SSDs. When making a storage plan, you need to know SSD performance.

SSDs have greater advantages over HDDs in terms of I/O models with hotspot data access and sensitivity to response latency. In particular, this is the case with random small I/O read models of database applications. SSDs do not have the spin latency that can be seen in HDDs. In bandwidth-sensitive applications, SSDs slightly outperform HDDs. In the tiered storage technology, SSDs are used to compose the high-performance tier required for high IOPS pressure.

If a performance problem occurs and the front end of the storage system is normal, check whether disks have reached the performance threshold. If this is the case (the disk usage level has reached almost 100%), then there is a restriction issue with the back-end disk performance as the IOPS and bandwidth cannot increase. To query disk utilization levels, use DeviceManager.

To ensure the reliability and full service life of a disk, keep usage below 70% as recommended. If the usage of most disks in a disk domain is greater than 90%, you are advised to add disks to the disk domain or migrate services to disks with better performance.