Running Docker periodically crashing on "CRITICAL: Failed to get speedtest.net configuration. Aborting"

[lrhenness]

Issue: The docker container aborts on failure to get Speettest configuration periodically from different servers.

Occurrence: Seemingly random. In my use case I have a list of about 8 servers across the US testing. The tool will do it's job over several iterations of those 8 servers. I will log back on some time later and discover the container crashed with the above log.

Cause: I suspect that for whatever reason, a temporary drop in connectivity between the program and the Speettest server will cause the failure.

Idea: Instead of the failure to obtain a config aborting the entire program, a failure to connect will initiate a retry/skip timer for that particular server, allowing the program to continue running having dropped a single test rather than all tests from the crash until next reboot.

Temporary Solution: Adding a docker --restart-always when starting the container

[techie2000]

Just checked my Influx logs for similar and can confirm that I am. I've never really noticed it as I run a fair number of dockers by default in --restart-always mode on my home rig!

  | 2020-09-09 15:53:01 | 2020-09-09 14:53:01,310 - CRITICAL: Failed to get speedtest.net configuration.  Aborting
-- | -- | --
  | 2020-09-09 15:50:50 | 2020-09-09 14:50:50,534 - CRITICAL: Failed to get speedtest.net configuration.  Aborting
  | 2020-09-09 15:47:34 | 2020-09-09 14:47:34,342 - CRITICAL: Failed to get speedtest.net configuration.  Aborting
  | 2020-09-09 15:45:23 | 2020-09-09 14:45:23,542 - CRITICAL: Failed to get speedtest.net configuration.  Aborting
  | 2020-09-09 15:43:12 | 2020-09-09 14:43:12,736 - CRITICAL: Failed to get speedtest.net configuration.  Aborting
  | 2020-09-09 15:41:01 | 2020-09-09 14:41:01,928 - CRITICAL: Failed to get speedtest.net configuration.  Aborting