Skip to main content

How Metric Collections Works

Metrics are collected every 1 minute. Monitoring data, such as read/write rate, read/write IOPs, and read/write latency, is available at 30-second granularity. Space-related data is collected at a polling frequency of 1 minute.

IO provides data at the VolumeGroup and StorageArray level based on volume information and relationships.

The Pure1 mirror replication metrics are mapped to the following IO metrics:

  • Avg Remote Latency Per Sync Write Operation

  • Avg Sync Replication Write Byte Rate

  • Avg Sync Replication Write IOPS

  • Avg Sync Replication Write IO Size

In the latest version of Pure Integration, we have added support for metric collection for Network Interfaces in form of Pure FC and Ethernet Ports.

This version of the Pure Integration adds metric collection for FlashBlade object store entities. Capacity metrics are collected for realms, object store accounts, and buckets. Read and write throughput, IOPS, and latency metrics are also collected for buckets. This collection is disabled by default; select Collect Object Store in an array’s configuration to enable it for that array.

The complete list of metrics for this integration is available in a separate guide.