Troubleshoot Guide#

This article contains troubleshooting steps for known issues. If you encounter any problems with the application not addressed here, check the GitHub Issues board. Feel free to file new tickets there (after learning about the guidelines for Contributing).

1. Seeing “No Data” in Grafana#

Issue#

Grafana panels show “No Data” even though the container/stack is running.

Reason#

The system date/time is incorrect on the device. If the system time is wrong, data timestamps fall outside Grafana’s query window.

Solution#

Check the date/time using the command below:

date

Set the correct date/time manually:

sudo date -s 'YYYY-MM-DD HH:MM:SS'   # Replace with your actual date and time

Set date/time from the internet:

sudo date -s "$(wget --method=HEAD -qSO- --max-redirect=0 google.com 2>&1 | sed -n 's/^ *Date: *//p')"

2. Influx – Data Being Deleted Beyond Retention Policy (RP)#

Issue#

  • Data appears to be deleted beyond the configured retention policy (RP).

  • InfluxDB 1.x deletes old data based on the retention policy duration and shard group duration.

Reason#

  • Data is grouped into shards.

  • Shards are deleted only when all data inside them is older than the RP.

  • For RPs ≤ 2 days, shard group duration = 1 hour.

  • InfluxDB always expires data at RP + shard duration.

Example:

For a 1-hour RP: - Data written at 00:00 goes into the shard covering 00:00–01:00. - The shard closes at 01:00. - InfluxDB deletes the shard only when everything inside it is past the RP → at 02:00.

So the effective expiration time is 1 hour RP + 1 hour shard duration = 2 hours.

Retention Policy

Shard Duration

Actual Expiry

1 hour

1 hour

2 hours

2 days

1 hour

2 days + 1 hr

30 days

24 hours

30 days + 24 hr

Solution#

  • Understand that this is normal and expected behavior in InfluxDB 1.x.

  • A 1-hour RP will always result in ~2 hours before deletion.

  • No configuration can force deletion exactly at the RP limit.

3. Time Series Analytics Microservice (Docker) – Takes Time to Start or Shows Python Packages Installing#

Issue#

The Time Series Analytics Microservice takes time to start or displays messages about Python packages being installed.

Reason#

UDF packages require several dependent packages to be installed during runtime, as specified under udfs/requirements.txt. Once these dependencies are installed, the Time Series Analytics microservice initializes and starts inferencing.

Solution#

No action required — wait for the time-series-analytics microservice to complete downloading the dependent packages and initialize Kapacitor to start inference.