Docker resource detection fails after system reboot

[litetex]

Component(s)

processor/resourcedetection, processor/resourcedetection/internal/docker

What happened?

Description

I'm running a container of this project with the resourcedetectionprocessor/docker.

The corresponding config looks like this:

processors:
  resourcedetection:
    detectors: [docker]

After a system restart the resource detector loses it's functionallity (the lables from the docker host are no longer picked up) and the following message shows up in the log:

warn    internal/resourcedetection.go:130       failed to detect resource       {"kind": "processor", "name": "resourcedetection", "pipeline": "logs", "error": "failed getting OS type: failed to fetch Docker OS type: Get \"http://%2Fvar%2Frun%2Fdocker.sock/v1.46/info\": context deadline exceeded"}

This is likely because docker is currently starting up.

Possible fixes:

• Retry resource detection for a limited amount of time (a few minutes?)
• Crash the container when this happens so that it can get restarted automatically due to the restart policy

It would be great to also make this configureable.

Collector version

v0.107.0

Environment information

Environment

Ubunut 24.04 LTS
Docker CE 27.1.2

OpenTelemetry Collector configuration

No response

Log output

No response

Additional context

No response

[github-actions]

Pinging code owners:

• processor/resourcedetection: @Aneurysm9 @dashpole

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[rogercoll]

Crash the container when this happens so that it can get restarted automatically due to the restart policy

Maybe a new error_mode configuration option like the transform processor would help, when set to propagate it should return any detector error and crash the collector.

[rogercoll]

Another viable solution would be what is being proposed here: #34876

[github-actions]

This issue has been inactive for 60 days. It will be closed in 60 days if there is no activity. To ping code owners by adding a component label, see Adding Labels via Comments, or if you are unsure of which component this issue relates to, please ping @open-telemetry/collector-contrib-triagers. If this issue is still relevant, please ping the code owners or leave a comment explaining why it is still relevant. Otherwise, please close it.

Pinging code owners:

• processor/resourcedetection: @Aneurysm9 @dashpole

See Adding Labels via Comments if you do not have permissions to add labels yourself.

[litetex]

explaining why it is still relevant

It's still not fixed so. This should be reason enough.