X-axis units as field value in series instead of time

Is it possible to plot the points on my graph so that the x-axis values represent a field value (an incremental identifier) instead of the timestamp for each point?

I am trying to graph the duration of events which occur on a non-uniform frequency, but I’d like the graph to show the points with a uniform distribution along the x-axis (as opposed to being spread out by the time between each event)

Thanks

Do you mean the X-Axis mode option where you can choose between Time (the default) and Series:

Close, but that lumps all of my points into one bar, whereas I still want to plot each data point separately.

I had hoped that using the Series X-Axis mode in addition to the ‘Format as Graph’ option in my query would do the trick, but it seems not.

I just came across this as well which seems to be the same thing I’m after: https://github.com/grafana/grafana/issues/7470

The histogram workaround sounds less than ideal though. Are there any other options?

You did not say which time series db you are using (is it ElasticSearch?). Can you not just to do a group by/aggregation?

Here is an example with ElasticSearch:

Here is an example with InfluxDB:

I am using InfluxDb.

Unfortunately the value I’d want to group by is a field (since its a UUID/guid, and I don’t want high series cardinality) and it seems that its only possible to group by tags.

Any other thoughts?

Thanks!

Any other thoughts?

Regarding? InfluxDB cannot only group by tags?

As far as I know InfluxDb can only group by tags and time intervals.

A request was made to group by fields (https://github.com/influxdata/influxdb/issues/7200) but this was closed in lieu of work to support high cardinality series (https://github.com/influxdata/influxdb/issues/7151). Once the high cardinality series support is in then I think I can tag my data points with the UUID, but until then I need to store it as a field.

So you want to group by UUID but InfluxDB cannot group the data that way? I’m not sure I see a way to solve that in Grafana :pensive: