Investigating the performance of storage resources

You can investigate the information that is collected about the storage resources that you add to IBM Storage Insights Pro. To identify performance issues, the information that is collected is shown in charts and tables.

About this task

Use the performance charts and tables to analyze, detect, and investigate issues that might affect your storage environment such as slow response times and workloads with high I/O rates.

Try it out! From the modern UI of IBM Storage Insights Pro, Go to Main menu, expand Insights, and click Performance.
Note: The modern UI currently supports fleet-level performance analysis for storage systems. Performance views on device details pages and individual component pages remain in the classic UI.
Get a head start on how to use the performance charts to detect and investigate performance anomalies:
Table 1. Actions and features of performance charts
Tiles Description
Performance Use breadcrumbs to navigate. Select the storage system type from the Device type list. The modern UI currently supports fleet-level performance charts for block storage systems only.
Storage systems View all available storage systems. By default, 6 systems are selected to simplify the view. You can select up to 10 storage systems. Click the expand icon to view maximum, minimum, and average values for selected metrics.
  • Use the search box to find storage systems.
  • Select IO rate or Response time from the list to view the metric with the storage systems.
  • Click the overflow menu beside the list to view and export cumulative performance for components of selected storage systems.
  • Use the overflow menu for each storage system to view its performance or add it to general groups.
Toolbar Select your time zone, customize the chart period by changing start and end dates, and choose the granularity (sample, hourly, or daily). Export performance data in CSV, PNG, or JPG format. Click the expand icon to view the chart in full screen.
Performance graph The graph shows storage system performance trends over time.
  • The y-axis shows the unit of measurement for each metric. The x-axis shows time.
  • Click Edit to select storage systems for the performance graph. Click Add metrics to add or change performance metrics. You can add up to 6 metrics. When you add multiple metrics, new graphs appear below existing graphs. Toggle the eye icon to show or hide charts.
  • Each graph displays one metric for up to 10 storage systems (default is 6). For example, selecting 2 metrics and 3 storage systems displays 2 graphs with 3 lines each. Each line represents a metric for one storage system.
  • Hover over points on a line to view performance information at that time.
  • Use the zoom bar or scroll the mouse wheel to zoom in or out of specific graph areas.
  • Each graph supports full-screen view, data refresh, zoom, table view, and image export.
  • Click Show details under each graph to view data in table format.

If you are in the classic UI of IBM Storage Insights Pro, go to Resources > Block Storage Systems and click Performance.

Note: The image that is provided is a mock-up of the Performance tab to illustrate the information that is provided and the actions that you can complete.
Performace tab
Get a head start on how to use the performance charts to detect and investigate performance anomalies:
Table 2. Actions and features of performance charts
Number Description
1

When you click the Performance tab, a highly granular chart is automatically generated that tracks the performance of key metrics, such as I/O rates and response times over the last 12 hours. By default, charts are generated for 5 of the resources.

To refine the information that is shown in the chart, you can complete these actions:
  • Click the resource in the Selected Resources pane to check the performance of a single resource.
  • Press Ctrl and Shift and then click the resources to compare the performance of two or more resources.
  • Click resources in the table under the chart to add them to the chart.
2 Highly granular charts at sample data resolution are shown by default for the last hour, 6 hours, 12 hours, and day. That is, performance is analyzed, measured, and reported at 5-minute intervals. To analyze performance at sample data resolution for longer periods, such as 1 or 2 weeks, reduce the number of resources that are selected. You can also use the calendar tool to customize the reporting period to a number of days instead of a week or a month.
3

To switch from chart to a table view of the performance metrics that are shown in the chart, click the Toggle button.

Table view of charts You can download the performance metadata that was used to generate the chart. Although the charts aren't included in the download, you can use the metadata to generate your own charts.
4 Select the time zone from the drop down.

You can select the time zone in which you want the performance metrics to be visible when you export it.
5 To export the performance metrics that are shown in the chart to a spreadsheet, click the Export button.
6

To open the chart in a new window, click the Open in new window button.

For example, you detect slow response times for one of the storage systems. You want to find out which of the storage system's internal resources is causing the slow response times. To investigate, you want to examine and compare the performance metrics that are shown at the storage system level with the performance metrics that are shown at the node, pool, or volumes levels. So, you complete these actions:
  1. Open a new window.
  2. In the Selected Resources pane in the new window, click the storage system.
  3. Click Actions and select the internal resource from the list.
Repeat this procedure to compare the performance metrics to detect the resources that are causing the slow response times.

Tip synchronize time in charts When you open a new window, you can click the Synchronize Time button to synchronize the selected time period with the time period of the first chart that you opened.
7 To add or change the performance metrics that are shown in the chart, click the Add Metrics button. You can add up to six metrics with two different metrical units to the chart.
8 You can customize the period that is examined in the chart by changing the start and end dates.
If applicable, three values are calculated for each of the performance metrics that are shown for the resources in the table:
  • The average value of the performance metric over the selected period.
  • The minimum value of the performance metric over the selected period.
  • The maximum value of the performance metric over the selected period.
Table 3. Actions and features of the performance table
Number Description
9 Use the filter to group the resources that you want to view in the table and the chart. For example, if you have a naming convention for types of storage systems, you can type the letters in the filter and press Enter. The chart and the table are updated to show the resources that you want to see.
10 Click Add columns to table to add information about the resource to the table. For example, you can add information about Coverage (%) to the table to measure the availability of performance metrics over the selected time period.
11
Explore the following actions:
  • Click Actions and select an internal resource. The resource opens in a new window so you can investigate which of the internal resources are overloaded or experiencing high response times.
  • Click Edit Table Metrics to change or add more performance metrics to the table. You can add as many performance metrics as you like to the table.
  • Click Create Report to share information by email about performance with your colleagues.