Downsample data with notebooks

Downsample data to reduce the overall disk usage as data collects over time. Learn how to create a notebook that downsamples data—walk through the following example to create a notebook that does the following:

  • Inputs InfluxDB Cloud demo data from the Website Monitoring Bucket
  • Downsamples data by aggregating data within windows of time with a Flux script
  • Outputs downsampled data to a bucket


Downsample data with a notebook

  1. Create a new notebook (see Create a notebook).
  2. Add a Metric Selector cell to select a bucket to query data from:
  • In the Choose a bucket dropdown list, select Website Monitoring Bucket.
  • Select the response_time field.
  • Select the mean function from the aggregate selector in the upper-right.
  1. Add a Flux Script cell with the following script to bring in data from the previous cell and downsample it:
  |> aggregateWindow(fn: mean, every: 1h)

If the every duration is longer the the total time range queried, aggregateWindow will only return one value.

4. Add an Output to Bucket cell, and then select the destination bucket for your downsampled data. 5. Add a Markdown cell to add a note to your team about what this notebook does. For example, the cell might say, “Downsample to one value per hour so we can do week-over-week performance. Sending data to downsample bucket.” 5. Click Preview in the upper left to verify that your notebook runs and preview the output. 6. Run your notebook:

  • Click Run from the Preview dropdown list to run the notebook and write to the output bucket a single time.
  • To write continuously, click Export as Task in the upper right corner of the Output to Bucket cell. For details about working with tasks, see Manage tasks.

Cloud or OSS?

InfluxDB OSS 2.0 now generally available!

InfluxDB OSS 2.0 is now generally available and ready for production use. See the InfluxDB OSS 2.0 release notes.

For information about upgrading to InfluxDB OSS 2.0, see: