Knowledge Base Administration Guide

Charts by Example

Here are a variety of basic charts you can generate with Simscope Metric Charts, showing both an example chart and the configuration used to render it.

Chart Configuration overview

Overview


Weekly Pass Rate, by Job Category

Note: this is the simplest chart configuration, to see how charts work.

This is a 52-week Line Chart of average Pass Rate among all jobs, grouped by Job Category:

  • compileblue
  • gen (ie generate) — green
  • sim (ie simulate) — orange

Cycles

Chart Configuration:

Chart FieldValue
MetricPass Rate
Chart TypeLine
X AxisWeekly
Group Series ByJob Category

Daily Pass Rate, by Component

This is a 30-day line chart showing average Pass Rate for 3 components:

Cycles

Chart Configuration:

Chart FieldValue
MetricPass Rate
Chart TypeLine
X AxisDaily
Group Series ByComponent
Compute(enter a regex)

Weekly Cycles, by Component

This uses a Stacked Bar chart, which enables seeing total weekly cycles as well as individual component weekly cycles.

  • Each component's cycles are a different color

Cycles

Chart Configuration:

Chart FieldValue
MetricCycles
Chart TypeStacked Bar
X AxisWeekly
Group Series ByComponent

Weekly Compute Time Percentage, by Component

This uses a Percent Area chart, which enables seeing visually which components are using the highest and lowest percentage of total compute time, across each week.

  • Percent Area charts scale the Y values to 100%, so you can view percentages easily.

Cycles

Chart Configuration:

Chart FieldValue
MetricCompute Weeks
Chart TypePercent Area
X AxisWeekly
Group Series ByComponent

CPS (cycles per second), by Regression

This is a Line Chart showing CPS for a group of regressions.

  • You can click on any point in the line, to drilldown to regression details at that point.

CPS

Chart Configuration:

Chart FieldValue
MetricCPS (Pass Only)
Chart TypeLine
X AxisRegression (Scatter)
Group Series ByRegression Name
Component(enter regex)

Regression CPS (cycles per second), by Build

Within a series of Regressions, you can group by their build configuration.

For example, this charts CPS grouped by:

  • 1-core (blue)
  • 2-core (green)
  • 4-core (orange)

CPS

Chart Configuration:

Chart FieldValue
MetricCPS (Pass Only)
Chart TypeLine
X AxisRegression (Scatter)
Group Series ByBuild
Component(enter a regex)
Regression Name(enter a regex)

Running Total Weekly Cycles, by Component

This uses an Area Running Total chart, which enables seeing cumulative weekly cycles across each component for the past 52 weeks (year).

This is also called a summation chart, or a cumulative chart.

  • Each component's cycles are a different color
  • You can also see the total cycles for the year, across all components, was 550 billion

Cycles

Chart Configuration:

Chart FieldValue
MetricCycles
Chart TypeStacked Bar
X AxisWeekly
Group Series ByComponent

Pie chart, by Signature Classification

Pie charts are useful at showing ratios or comparisons of data.

This example shows the percentage of failures for the past 30 days, grouped by Signature Classification.

Pie

Chart Configuration:

Chart FieldValue
Metric#Fail
Chart TypePie
X AxisDaily
Group Series BySignature Classification

Regression Coverage Chart (using metadata)

If you are storing coverage values inside of Regression Metadata, you can create coverage charts with Simscope.

This example plots the metadata func_coverage and line_coverage, for a 100-day period.

Coverage

Chart Configuration:

Chart FieldValue
MetricRegression Metric Value
Chart TypeLine
X AxisRegression (Scatter Plot)
Group Series ByRegression Metric Value
Component(enter a regex)
Regression Name(enter a regex)
Regr Metadata(optional: enter a specific coverage series name)

Daily Regression Pass Rate Heatmap

Heatmap Charts are useful when visualizing many series simultaneously.

  • As opposed to line charts, which can become cluttered if there are more then 3 or 4 series in a single chart.

This is a 30-day Heatmap Chart of Pass Rate, grouped by regression name.

  • Note that when charting Pass Rate, the Heatmap chart will colorize the values (ie green, yellow, red) to make it easy to see patterns visually.

Heatmap

Chart Configuration:

Chart FieldValue
MetricPass Rate
Chart TypeHeatmap
X AxisDaily
Group Series ByRegression Name

Daily Pass Rate by Build (Heatmap)

This is a 10-day Heatmap Chart of Pass Rate, grouped by Build.

  • This lets you easily visualize the daily health of each build.

Heatmap

Chart Configuration:

Chart FieldValue
MetricPass Rate
Chart TypeHeatmap
X AxisDaily
Group Series ByBuild

Triage Rates

Simscope can chart Triage and Assignment rates (useful in line and heatmap/datatable charts).

For example, this chart shows a 14 day chart of the percentage of daily errors being triaged, grouped by Component.

Triage

Chart Configuration:

Chart FieldValue
Metric% Jobs Triaged
Chart TypeLine
X AxisDaily
Group Series ByComponent

Advanced: Regression Coverage Heatmap

If you are storing coverage values inside of Regression Metadata, you can create coverage charts with Simscope.

This example plots func coverage for all weekly regressions, across multiple components, for a 30-day period:

  • X-axis is grouped by model version (ie SHA).
    • For example, you can see multiple regression coverage results for checkin b2ffd (see the red arrow)

Coverage

Chart Configuration:

Chart FieldValue
MetricRegression Metric Value
Chart TypeHeatmap
X AxisModel Version
Group Series ByRegression Name
Regr Metadata(enter a specific coverage series name)
Component(enter a regex)
Regression Name(enter a regex)

Advanced: Regression Coverage by Model

If you are storing coverage values inside of Regression Metadata, you can create coverage charts with Simscope.

This example plots coverage values for a specific model version (ie checkin):

  • Model Version (SHA) b2ffd
  • Notice you can see both func_coverage and line_coverage for all regressions in one chart

Coverage

Chart Configuration:

Chart FieldValue
MetricRegression Metric Value
Chart TypeHeatmap
X AxisRegression Metric Name
Group Series ByRegression Name
Model Low(enter a model version/SHA)