core: Dramatic increase in the number of "Error doing job: Task was destroyed but it is pending!" errors

The problem

I started seeing periodic Error doing job: Task was destroyed but it is pending! messages in 0.114.x, but now I see them pretty regularly with 0.115.0b0, and from different components. I’ll start disabling my custom components, but wanted to get this in in case others are seeing something similar.

Environment

Home Assistant Core release with the issue: 0.115.0b0
Last working Home Assistant Core release (if known): 0.113
Operating environment (OS/Container/Supervised/Core): OS
Integration causing this issue: Various - recorder, MainThread
Link to integration documentation on our website: N/A

Problem-relevant `configuration.yaml`

Without knowing what’s involved in these errors, it’s hard to pick out what might be relevant from configuration.yaml.

Traceback/Error logs

Lots of these logs. I see “recorder” involved (I’m using the MariaDB addon, which is otherwise working fine), MainThread, and Thread-3 in here. There’s no obvious context around the messages to indicate what’s causing them. I can try turning debug on later today.

2020-09-08 09:00:18 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:08:39 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:09:09 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:11:09 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:12:39 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:13:29 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:13:39 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:13:59 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:19:59 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:20:39 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:20:49 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:25:29 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:30:30 INFO (MainThread) [homeassistant.components.automation.keep_water_hot] Keep Water Hot: Running automation actions
2020-09-08 09:30:30 INFO (MainThread) [homeassistant.components.automation.keep_water_hot] Keep Water Hot: Executing step call service
2020-09-08 09:31:59 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:34:29 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:36:39 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:39:59 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:40:49 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:44:29 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:45:30 INFO (MainThread) [homeassistant.components.automation.keep_water_hot] Keep Water Hot: Running automation actions
2020-09-08 09:45:30 INFO (MainThread) [homeassistant.components.automation.keep_water_hot] Keep Water Hot: Executing step call service
2020-09-08 09:47:09 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:50:09 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:51:49 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 09:52:49 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:00:39 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:04:39 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:06:09 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:07:59 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:08:59 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
sequence
2020-09-08 10:12:29 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:12:39 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:13:29 ERROR (Thread-3) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:14:49 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:17:29 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:19:29 ERROR (Recorder) [homeassistant] Error doing job: Task was destroyed but it is pending!
2020-09-08 10:19:59 ERROR (MainThread) [homeassistant] Error doing job: Task was destroyed but it is pending!

Additional information

About this issue

Original URL
State: closed
Created 4 years ago
Comments: 25 (22 by maintainers)

Most upvoted comments

I’ll continue to try to narrow it down. I ruled out all of my custom components as the cause last night.

It would be really nice if there was a “disable” option for integrations, either in Configuration->Integrations or, for a more complete list, in Configuration->Info next to the issues/documentation links. It used to be simple to comment things out in configuration.yaml, but the UI-based config flow often requires a time-consuming process of removing and re-configuring an integration just for a quick test. The easiest alternative that I’ve found so far is to edit core.config_entries and add a “-1” or similar to the domain for the integrations that I want to disable, but that’s pretty hacky. There’s a feature request for this somewhere in the forum that I can’t find right now.

rccoleman on Sep 11, 2020

That seems to be easier said than done. I commented out my recorder: definition, default_config:, and everything that I thought used it and it’s still being setup by something. This is what I’m using instead of default_config: now:

# default_config:
automation: !include automations.yaml
cloud: !include cloud.yaml
config:
frontend:
  themes: !include_dir_merge_named themes
# history:
input_boolean: !include input_boolean.yaml
input_number: !include input_number.yaml
# logbook:
# map:
mobile_app:
person:
scene: !include scenes.yaml
script: !include scripts.yaml
ssdp:
sun:
system_health:
updater:
zeroconf:

I’m still seeing the errors after restarting with that, both a few during startup and then periodically thereafter.

I have a suspicion that it’s related to AppDaemon and some of the automations that interact with entities manipulated by both, but I haven’t been able to confirm it definitively. I’ll keep poking around.

rccoleman on Sep 9, 2020