A bar chart compares values across categories using horizontal or vertical bars. Each bar represents a category, and bar length represents the value.
Use the Plot panel to choose axes, grouping, and display settings.
Use these options to control how the bar chart renders.
X-axis
Field used for the X axis.
Use timestamp for time-based bars. Use a label field for categorical bars (for example service, pod, host).
timestamp
service
pod
host
Y-axis
Numeric field to plot.
Common choices: value, cpu_usage, latency, error_rate.
value
cpu_usage
latency
error_rate
Group by
Split into multiple bar series per group.
For example pod creates one series per pod. Leave empty for a single series.
Y-axis label
Display label for the Y axis.
Cosmetic only. Example: “CPU Usage (%)”.
Y-axis scale
How values are scaled on the Y axis.
Linear for most data. Logarithmic for wide ranges.
Reference line
A horizontal threshold line.
Example: 80 for an SLA/limit line.
80
Legend
Show/hide legend.
Turn off for a single series or tight layouts.
Calculations
Summary overlays on the plotted data.
Min, Max, Avg, Count, Sum, P50, P90, P95, P99.
Smooth graph
Smooths the rendered curve/series.
Useful for long time ranges. Avoid for step-like data.
Custom color
Set fixed colors for series.
Helps keep colors consistent across dashboards.
Apply
Applies changes to the widget.
Changes do not render until you click Apply.
Use a bar chart when you want to:
Compare values across categories at a point in time
Rank items (top N / bottom N)
Compare groups side-by-side (grouped bars)
Show distribution across discrete buckets
Typical use cases:
Requests by service / endpoint
Errors by status code
Cost by team / namespace
Latency by region
Avoid a bar chart when:
You need trends over time across many points (use Line/Area)
You need part-to-whole as percentages (use Pie/Disk)
You need detailed row-level inspection (use Table/List/Details)
Last updated 5 months ago
Was this helpful?
