Master Grafana Time Series Panel: From Basics to Advanced Configuration

Why Time Series Is Grafana’s “Soul”?

Stat panels show the current state, but without a time dimension you cannot predict future problems. The Time Series panel visualizes metric values over time, revealing trends, patterns, cycles, and anomalies, turning blind monitoring into a predictive tool.

Key Questions Without a Time Dimension

Is a 90% CPU spike sudden or gradual?

Has a similar peak occurred before, and how long did it last?

Do memory and disk I/O change together with the peak?

Is the trend periodic (e.g., daily rush hour) or a one‑off event?

Configure Time Series Visualization in the Browser

Select the Time series visualization type in Grafana’s panel editor.

PromQL Settings

Metrics browser : write PromQL expressions to fetch the desired data.

Option.Legend : format the legend fields for clearer display.

Panel Options

Title : set a descriptive panel name, e.g., “Host Memory”.

Tooltip : choose multi mode to show values for all series on hover; set value sort order to descending.

Legend : toggle visibility, choose display mode (table), placement (right), and which calculations to show (max, mean, last non‑null).

Axis : set time zone to browser , choose Y‑axis placement (auto), and label the axis (e.g., “Host Memory Usage”).

Graph Styles : smooth line interpolation, line width 2, fill opacity 25, gradient mode opacity .

Standard Options : set unit to bytes and choose the Classic color palette.

JSON Configuration Example

{
  "datasource": {"uid": "${datasource}"},
  "fieldConfig": {
    "defaults": {
      "color": {"mode": "palette-classic"},
      "custom": {
        "axisLabel": "主机内存使用情况",
        "axisPlacement": "auto",
        "fillOpacity": 25,
        "gradientMode": "opacity",
        "lineInterpolation": "smooth",
        "lineWidth": 2
      },
      "unit": "bytes"
    },
    "overrides": [{
      "matcher": {"id": "byRegexp", "options": "/.*使用率/"},
      "properties": [
        {"id": "custom.axisPlacement", "value": "right"},
        {"id": "custom.axisLabel", "value": "主机内存使用率"},
        {"id": "unit", "value": "percentunit"}
      ]
    }]
  },
  "gridPos": {"h":9,"w":24,"x":0,"y":0},
  "options": {
    "legend": {
      "calcs": ["max","mean","lastNotNull"],
      "displayMode": "table",
      "placement": "right",
      "showLegend": true
    },
    "timezone": ["browser"],
    "tooltip": {"mode": "multi", "sort": "desc"}
  },
  "targets": [
    {"editorMode":"code","expr":"node_memory_MemTotal_bytes{instance=~\"$instance\"} - node_memory_MemFree_bytes{instance=~\"$instance\"} - (node_memory_Cached_bytes{instance=~\"$instance\"} + node_memory_Buffers_bytes{instance=~\"$instance\"} + node_memory_SReclaimable_bytes{instance=~\"$instance\"})","legendFormat":"{{instance}}","refId":"A"},
    {"editorMode":"code","expr":"(node_memory_MemTotal_bytes{instance=~\"$instance\"} - node_memory_MemFree_bytes{instance=~\"$instance\"} - (node_memory_Cached_bytes{instance=~\"$instance\"} + node_memory_Buffers_bytes{instance=~\"$instance\"} + node_memory_SReclaimable_bytes{instance=~\"$instance\"})) / node_memory_MemTotal_bytes{instance=~\"$instance\"}","legendFormat":"{{instance}}使用率","refId":"B"}
  ],
  "title": "主机内存",
  "type": "timeseries"
}

Conclusion

The Time Series panel is Grafana’s “eye of trends”. By mastering data handling, visual configuration, and multi‑metric correlation, you can shift from reactive “fire‑fighting” to proactive “fire‑prevention”.