Spark Charts in List Designer/Viewer do not show variances in the data properly

Products

• DX Unified Infrastructure Management (Nimsoft / UIM)

Issue/Introduction

I have created a List View with a Spark chart that shows a QoS metric for a long time range (e.g. last 30 days)

Even though I know there have been some spikes in the data, all the bars look flat.

Environment

UIM 23.4.3  (see Additional Information below for prior releases)

Cause

This is related to the automatic scaling.

This occurs when the data has a large range (like 0-100 for CPU data) but the spikes are small (e.g. 10% or less.)

The bars appear relative to the entire data range, so 10% CPU only takes up 10% of the chart space.   At these scales it is difficult to differentiate between 5% and 10%.

Resolution

To resolve this, you can set a custom scale.  Uncheck "Automatic Scale" and then choose an appropriate scale.

For example, if you know the CPU value rarely goes above 20%, you can set a custom scale of 0 to 20.

This way, a 5% value would take up a quarter of the chart, and a 10% value would take up half of the chart, making it much easier to differentiate.

 

Additional Information

Custom scaling currently does not function properly in DX UIM 23.4.2 and prior but this will be addressed in DX UIM 23.4.3. (See related KB)

For now, there is no workaround in DX UIM 23.4.2 and prior, but when 23.4.3 is released, you can upgrade to that version and then set a custom scale.

ETA for 23.4.3 is the end of 2025.