Skip to main content
Pure Technical Services

KB: FlashArray API Response Metric Object Definitions

Currently viewing public documentation. Please login to access the full scope of documentation.

KP_Ext_Announcement.png

Question

How do I find out what the definitions are for the performance metric object names that are included in the API call response from the FlashArray API?

Applies to

FlashArray API, PowerShell SDK, Python SDK, py-pure-client

Answer

The definitions are explained in the Python py-pure-client documentation located here - py-pure-client ReadTheDocs . Specifically, refer to the section on "Models".

As an example, when querying the performance metrics for the FlashArray "array" class by:

  • API call example - GET http://flasharray/api/2.2/arrays/performance
  • PowerShell SDK2 cmdlet example - Get-Pfa2ArrayPerformance

The response definitions are shown as follows under this section of the documentation:

  • bytes_per_mirrored_write (int) – The average I/O size per mirrored write. Measured in bytes.
  • bytes_per_op (int) – The average I/O size for both read and write (all) operations.
  • bytes_per_read (int) – The average I/O size per read. Measured in bytes.
  • bytes_per_write (int) – The average I/O size per write. Measured in bytes.
  • mirrored_write_bytes_per_sec (int) – The number of mirrored bytes written per second.
  • mirrored_writes_per_sec (int) – The number of mirrored writes per second.
  • qos_rate_limit_usec_per_mirrored_write_op (int) – The average time it takes the array to process a mirrored I/O write request. Measured in microseconds.
  • qos_rate_limit_usec_per_read_op (int) – The average time spent waiting due to QoS rate limiting for a read request. Measured in microseconds.
  • qos_rate_limit_usec_per_write_op (int) – The average time that a write I/O request spends waiting as a result of the volume reaching its QoS bandwidth limit. Measured in microseconds.
  • queue_usec_per_mirrored_write_op (int) – The average time that a mirrored write I/O request spends in the array waiting to be served. Measured in microseconds.
  • queue_usec_per_read_op (int) – The average time that a read I/O request spends in the array waiting to be served. Measured in microseconds.
  • queue_usec_per_write_op (int) – The average time that a write I/O request spends in the array waiting to be served. Measured in microseconds.
  • read_bytes_per_sec (int) – The number of bytes read per second.
  • reads_per_sec (int) – The number of read requests processed per second.
  • san_usec_per_mirrored_write_op (int) – The average time required to transfer data from the initiator to the array for a mirrored write request. Measured in microseconds.
  • san_usec_per_read_op (int) – The average time required to transfer data from the array to the initiator for a read request. Measured in microseconds.
  • san_usec_per_write_op (int) – The average time required to transfer data from the initiator to the array for a write request. Measured in microseconds.
  • service_usec_per_mirrored_write_op (int) – The average time required for the array to service a mirrored write request. Measured in microseconds.
  • service_usec_per_read_op (int) – The average time required for the array to service a read request. Measured in microseconds.
  • service_usec_per_write_op (int) – The average time required for the array to service a write request. Measured in microseconds.
  • time (int) – The time when the sample performance data was taken. Measured in milliseconds since the UNIX epoch.
  • usec_per_mirrored_write_op (int) – The average time it takes the array to process a mirrored I/O write request. Measured in microseconds. The average time does not include SAN time, queue time, or QoS rate limit time.
  • usec_per_read_op (int) – The average time it takes the array to process an I/O read request. Measured in microseconds. The average time does not include SAN time, queue time, or QoS rate limit time.
  • usec_per_write_op (int) – The average time it takes the array to process an I/O write request. Measured in microseconds. The average time does not include SAN time, queue time, or QoS rate limit time.
  • write_bytes_per_sec (int) – The number of bytes written per second.
  • writes_per_sec (int) – The number of write requests processed per second.
  • service_usec_per_read_op_cache_reduction (float) – The percentage reduction in service_usec_per_read_op due to data cache hits. For example, a value of 0.25 indicates that the value of service_usec_per_read_op is 25% lower than it would have been without any data cache hits.
  • id (str) – A globally unique, system-generated ID. The ID cannot be modified and cannot refer to another resource.
  • name (str) – A locally unique, system-generated name. The name cannot be modified.
  • queue_depth (int) – Deprecated. The number displayed here may not be accurate and in later versions of the product this field will always display null. Instead, use queue_usec_per_mirrored_write_opqueue_usec_per_read_op, and queue_usec_per_write_op fields to measure IO queuing.
  • local_queue_usec_per_op (int) – Average local queue time for both read and write operations, measured in microseconds.
  • usec_per_other_op (int) – The average time it takes the array to process an I/O other request, measured in microseconds. The average time does not include SAN time, queue time, or QoS rate limit time.
  • others_per_sec (int) – The number of other requests processed per second.