Grafana v7.4.3 Variable "returned too many points in a series"

Hello everybody,
I created this query to get the data from influxdb 2.0:

import “influxdata/influxdb/schema”
schema.measurementTagValues(
bucket: “MyBucket”,
measurement: “MyMeasurement”,
tag: “MyTag”
)

Grafana does not return all values!
Is there a maximum number of the number of values I can display in the variables?
In some cases I may have more than 3000 probes to view.

if I report the same query in the tables, I see all the values.
In some cases, when there are too many values, grafana returns the following error: “returned too many points in a series: 2976”

Is there any solution?
Thank you

That’s a logical problem: how to visualize X datapoints of timeseries if you have only Y pixels (usually width of the panel). If X < Y then it is fine. But what if X > Y, e.g. 3000 datapoints and you have only 300 pixels. But what if user select last year time range where query can return 1M+ datapoints?

Solution is time aggregation, so instead of full data resolution query returns only aggregated values per selected time period (e.g. for year time range query returns monthly values → so instead of 1M+ datapoints only 12 datapoints is returned). Grafana InfluxDB doc has examples how to do that with Grafana macros:

Please focus on |> aggregateWindow(every: v.windowPeriod, fn: mean)

Ok,
but every 15 minutes I have more than 3000 different probes that give data.
In grafana is it possible to use a query that returns me more than 3000 distinct records?
If not possible, I will try to divide the probes into several subgroups.
Thanks so much

Yes, it is possible. But everything has own limits and default Grafana limits are reasonable - max data points = width of the panel:

Nobody is stopping you to use own limit for Max data point

What will you do if you want to show 1M datapoints (e.g. last year of your data)? You will be waiting a few seconds until all data are transferred, then a few seconds until data are rendered, then maybe your browser will be consuming so much memory that your browser/OS will be freezing. Then you will start complaining that Grafana is slow. And that’s only because you have not poor query design. I would avoid any further issues with proper query design now.

1 Like