Saturation

The saturation chart provides information about the:

  • Saturation levels of the storage system during the selected time interval.

  • Periodicity in the change of saturation levels.

  • Saturation insights.

The saturation chart in the Analytics view indicates the saturation levels of the storage system that are measured using data samples from the selected time interval.

Calculating saturation level: Saturation level is calculated based on the following:

  • Type of workload that is running on the storage system: The workload running on a storage system depends on the request size and the percentage of read and write throughput that is serviced by the system.

  • Virtual volume types.

  • Configuration of the storage system. The configuration depends on the following:

    • Model

    • Number of nodes

    • Drive type

    • Port type

    • Number of chassis

    • Volume type (compression, dedupe, and thin)

The relevant metrics are measured at a 5 minute interval and the preferred throughput (number of inputs or outputs per second) is calculated. The preferred throughput the storage system is able to deliver within a reasonable response time is defined as the preferred throughput. The preferred throughput is measured against the actual throughput that is being serviced by the storage system. The preferred throughput varies with the configuration and the type of workload that is running on the storage system. The ratio is represented as percentage in the line chart.

NOTE:

Saturation levels can exceed 100%, if you have a storage system with lesser than the optimal number of drives. The preferred throughput is an indicator of the optimal threshold. The storage system is expected to perform appropriately when the saturation is within the optimal threshold. However, the system could still deliver more inputs/outputs per second beyond this level but the latency would be high.

For more information on saturation insights, see HPE SSMC User Guide.