Grafana Dashboard Load Calculator
Estimate query load a Grafana dashboard places on its underlying data sources.
About Grafana Dashboard Load Calculator
Each panel issues its own query on every auto-refresh tick: queries per viewer per second = panels / refresh_interval. Multiple concurrent viewers multiply that rate directly since each browser session refreshes independently by default. Concurrent query load (rate × average query duration) approximates how many queries are in flight on the data source at any instant — this is the number that actually determines datasource CPU/memory pressure, not the raw query rate alone.
Formula
load_time_sec = panels × avg_query_time_sec × (1 + concurrent_users × contention_factor)
- panels
- Number of panels on the dashboard
- avg_query_time_sec
- Average query execution time per panel (seconds)
- concurrent_users
- Users loading the dashboard simultaneously
- contention_factor
- Backend contention multiplier per additional user
Worked example
With the default values below, Grafana Dashboard Load Calculator returns Total Query Rate of 4.00. Change any input above and every figure here is recalculated in your browser.
Inputs used
- Panelsaccepts 1–500 panels
- 24 panels
- Auto-Refresh Intervalaccepts 1–3,600 seconds
- 30 seconds
- Average Query Durationaccepts 1–60,000 ms
- 250 ms
- Concurrent Viewersaccepts 1–100,000 viewers
- 5 viewers
Results
- Total Query Rate
- 4.00 queries/sec
- Concurrent Query Load on Data Source
- 1.00
- Estimated Data Source Memory Impact
- 2.0 MB
Step by step
- 1Queries per refresh cycle= 24 queries
- 2Queries/sec per viewer: panels ÷ refresh interval= 0.800/s
- 3Total queries/sec across viewers= 4.00/s
- 4Concurrent query load: rate × avg duration= 1.00 concurrent queries
Assumptions and limits
- The result is derived only from the 4 inputs listed above. Anything not among them is not part of the calculation.
- In particular, this calculator does not take inflation, tax, fees or charges or compounding frequency into account.
- Figures are rounded for display to 2 decimal places, so totals built from several rounded parts can differ slightly from the unrounded arithmetic.
How these calculators are built and tested is described on the Methodology page. If a figure looks wrong, please report it.
Frequently Asked Questions
Does Grafana cache query results across viewers?
Not by default for live dashboards — each browser session independently issues its own queries on refresh. Grafana Enterprise/Cloud query caching, or a caching proxy in front of the data source, can substantially cut this load for popular dashboards.
How much does a shorter refresh interval cost?
Query rate is inversely proportional to refresh interval, so halving the interval doubles the query rate — for dashboards with many viewers, this compounds quickly and is often the first thing worth increasing when a data source is under load.
Should every panel really run its own query?
Not necessarily — panels querying the same underlying series with different visualizations can sometimes share a single query via Grafana's mixed data sources or query result transformations, cutting effective panel-to-query ratio below 1:1.
What's a safe refresh interval for a busy dashboard?
For dashboards viewed by many people simultaneously, 30-60 seconds is a common balance; sub-10-second refresh is usually reserved for small, single-viewer, incident-response dashboards where up-to-the-second data matters more than aggregate load.