Overview
Monitor your Pure Storage FlashArray fleet using this Dynatrace Extension 2.0 and the Pure Storage FlashArray OpenMetrics exporter.
The integration can provide performance data at the array, host, volume and pod level, as well as high-level capacity and configuration information.
Monitor multiple FlashArrays and aggregate these into a single dashboard, or group them together by customer defined environment.
This integration requires the following:
The Pure Storage OpenMetrics exporter is installed and running in a containerized environment. Refer to the GitHub repo for installation instructions. https://github.com/PureStorage-OpenConnect/pure-fa-openmetrics-exporter
This extension provides 2x integrated dashboards:
Pure Storage Overview (OpenMetrics) Dashboard
Monitor your fleet of FlashArrays in one simple view, displaying
- Purity OS Version
- Alerts & Criticality
- Data Reduction Ratio
- Capacity Utilization
- Data Throughput
- IOPS by type
Pure Storage Entity Overview Dashboard
Drill down into individual FlashArray configurations and components.
Feature Sets
Explore and select metrics from the following feature sets:
- default (array details)
- Alerts
- Directories
- Hosts
- HW
- Network Interfaces
- Pods
- Volumes
Get started
To enable this extension, Dynatrace will need to connect to the Pure Storage OpenMetrics exporter. This exporter receives requests for exposed metric endpoints and sends the request to the FlashArray via the REST API client. The REST results are interpreted and sent back to Dynatrace in the OpenMetrics standard format for ingestion.
Please refer to Pure Storage OpenConnect GitHub repository for more information and the latest packages.
https://github.com/PureStorage-OpenConnect/pure-fa-openmetrics-exporter
Check the compatibility section for latest information on specific Dynatrace prerequisites.
Details
- Providing the Pure Storage FlashArray OpenMetrics Exporter has been deployed then the this extension can be configured.
- At the top of this page, use the green button 'Add to environment'.
- In the
Configuration tab of the extension under the Versions section select Add monitoring configuration.
- Monitoring source: Select
Monitor remotely tab.
- ActiveGates: Select the appropriate ActiveGate group and move to the
Next step.
-
Prometheus: Add configuration entries for each metric endpoint to be scraped. Refrain from scraping all metrics using /metrics as these are expensive queries in large environments.
- Select 'Add endpoint' button.
- URL:
http://pure-ome-address:9490/metrics/array?endpoint=arrayname01
- Authentication scheme: No authentication
- Skip ssl verification: Enabled
-
Finaize configuration: Check config and select required metric feature sets to be ingested into Dynatrace and select Activate
- Logon to the Dynatrace ActiveGate which has access to the OpenMetrics Exporter and restart the
extensionsmodule service to download and enable the configuration.
# systemctl restart extensionsmodule
- After a few minutes the configuration will appear with status
Ok and metrics will be ingesting.
Metric Endpoints & Scrape Intervals
For performance metrics we wish to scrape as frequently as possible to
By default this extension scrape endpoints at predefined intervals as follows:
| Feature Set | Device Metric Endpoint | Scrape Interval |
|---|
| default | /metrics/array | 1 min |
| Alerts | /metrics/array | 5 min |
| Directories | /metrics/directories | 5 min |
| Hosts | /metrics/hosts | 1 min |
| HW | /metrics/array | 5 min |
| Pods | /metrics/pods | 1 min |
| Volumes | /metrics/volumes | 1 min |
To customize these scrape intervals, contact your Dynatrace representative to configure a custom extension.
DDU Cost Calculation Per Array :
- Information Needed:
- Number of Hosts
- Number of Pods
- Number of Volumes
- Number of NICs
- Number of Hardware Components
- Number of Temp Sensors
- Number of Power Supplies
- Number of Directories(Filesystems)
(45 + (39 * Hosts) + (56 * Pods) + (38 * Volumes)) * 525.6)
+
((10 * NICs) + Total Hardware Components + Temp Sensors + Power Supplies + (26 * Directories) + 2) * 105.12)
= Total DDUs/Year for the one FlashArray
The first piece of the equation (45 + (39 * Hosts) + (56 * Pods) + (38 * Volumes)) * 525.6) are for metric polled every 1 minute
The second piece of the equation ((10 * NICs) + Total Hardware Components + Temp Sensors + Power Supplies + (26 * Directories) + 2) * 105.12) are for metrics polled every 5 minutes.
Use cases
This extension would be super useful for empowered Pure Storage Administrators and Site Reliability Engineers to capture metrics for FlashArray to utilize the awesome overview dashboard displaying high level fleet statistics and configuration.
Users will be able to:
- View high level array configuration
- Monitor performance metrics and configure alerts
- Correlate FlashArray metrics with other platforms
Compatibility information
Purity
Pure Storage FlashArray from Purity version 6.1.0 onwards. Some metrics may have been exposed in later versions of Purity REST API 2.x and Pure Storage OpenMetrics exporter.
Pure Storage OpenMetrics Exporter Configuration
Dynatrace requires secure connections when using any kind of authentication. Currently the Pure Storage OpenMetrics Exporter must be configured with:
- http with tokens file and no authentication
