Skip to content
Tutorials Center
My learning

Topics

Lesson 16 of 37 ยท Panels and Visualizations

The Time Series Panel: Lines, Bars, Points and Axes

Master the Grafana Time series panel: draw styles, line interpolation, fill, stacking, axes and soft limits, legends, tooltips, null handling, annotations and series overrides.

The Time series panel is the workhorse of Grafana: line graphs, area charts, bar charts and point plots of values over time. Almost every dashboard has several. Its option list is long, but it divides into a few logical groups.

When to use it

Use a time series panel whenever the question includes "over time": traffic, latency, memory, temperature, signups per hour. If you only need the current value use Stat or Gauge; for exact numbers use a Table.

Graph styles

Under Graph styles:

  • Style: Lines, Bars or Points.
  • Line interpolation: Linear, Smooth, Step before, Step after. Use step for values that change at discrete moments (a status, a configured limit); smooth lines can invent peaks that never happened.
  • Line width, fill opacity, gradient mode: fill turns a line into an area.
  • Line style: solid, dashed or dotted.
  • Show points: automatic, always or never.
  • Connect null values / Disconnect values: controls gaps. Never shows a gap when data is missing (honest); Always joins across the hole; Threshold joins only short gaps.
  • Stack series: Normal stacks values (total adds up), 100% shows share. Only stack values that can be meaningfully summed.

Axes

  • Placement: left, right or hidden; a series can use the right axis via an override.
  • Label: axis text such as "requests / s".
  • Scale: linear, log (base 2 or 10), symlog.
  • Soft min / soft max: suggest a range without clipping data that falls outside it (unlike hard min/max). Setting soft min = 0 keeps a latency graph from exaggerating small changes.
  • Centered zero, Show grid lines, Width.

Legend

Mode: List, Table (with columns) or hidden. Placement: bottom or right. Add calculations such as Last, Mean, Max, Min to the legend table so people can read numbers without hovering. Clicking a legend entry isolates that series; Ctrl/Cmd+click toggles several.

Tooltip

Modes: Single (nearest series), All (every series at the cursor time) or Hidden. Sort order matters when many series exist. All with descending order puts the biggest contributor first.

Thresholds as lines or regions

Under Thresholds you define steps (base green, 80 orange, 95 red). In the Show thresholds option choose lines, filled regions or both. This overlays limits on the graph, so a reader sees at once whether the value crosses a limit.

Annotations

Annotations from the dashboard's built-in query, or from data sources (deploy events from SQL, alerts), appear as vertical markers with hover text. Add one manually with Ctrl/Cmd+click on the graph.

Overrides

Overrides adjust specific fields: Fields with name, name matching regex, type, or returned by query. A classic use: draw "limit" as a dashed red line, and put "temperature" on a right axis with a different unit.

Stand-in: why interpolation matters

This stand-in turns step data into linear and step-after points to show how a smooth line can mislead for state-like data.

JS Stand-in: linear versus step interpolation

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Stand-in: null handling and stacking

JS Stand-in: gaps and stacked totals

Press Run (or Ctrl + Enter) to see the console output here.

No output.

Reading a good graph

  • Start with the title and unit, then the legend, then the shape.
  • Look for level shifts, spikes, periodicity (daily cycles) and gaps.
  • Compare with the same time yesterday using Time shift in query options.
  • Remember that zoom changes aggregation (see time ranges).

Performance

Too many series (hundreds) make graphs slow and unreadable. Aggregate with sum by or topk, or use a table or heatmap instead. Avoid showing points for dense data.

Check your understanding

Quick check ยท 4 questions

of 4 answered You scored / 4 ยท perfect!

  1. 1. Which interpolation best represents a value that changes at discrete moments, like replica count?

    Steps avoid inventing values between samples.
  2. 2. What is the difference between soft min and hard min?

    Soft limits adapt to data outside the range; hard limits fix the axis.
  3. 3. Which tooltip mode shows every series at the cursor's time?

    All lists every series; Single shows only the nearest one.
  4. 4. How can you draw a limit line as dashed red on the same panel?

    Overrides target fields and set color, dash and axis options.

Summary

  • Time series shows lines, bars and points over time.
  • Use step interpolation for state-like data and avoid inventing values.
  • Axes, soft limits, legends with calculations and tooltips improve readability.
  • Thresholds, annotations and overrides add context.
  • Reduce series counts before graphing.

Next: Stat, Gauge and Bar Gauge Panels