High cpu usage on zlinux nodes when collecting data using prometheus receiver

[phantonrider]

Component(s)

receiver/prometheus

What happened?

Description

When we are using prometheus receiver to collect metrics and the endpoint is down, the endpoints are local but the servers run active/passive, the opentelemetry collector will start to use more and more cpu until the it reaches 100%. When we disable the pipeline for this and leave the others running its fine and if the endpoint is available then its fine.

Only on zlinux and the server is a single core.

Steps to Reproduce

Setup the collector to collect from a endpoint that is not available on a single core machine.

Expected Result

Error message generated that the endpoint isnt available
CPU usage to remain low

Actual Result

Error message generated that the endpoint isnt available
CPU usage increases over time, within 24 hours

Collector version

0.116.0

Environment information

Environment

OS: zlinux

OpenTelemetry Collector configuration

unable to provide

Log output

No response

Additional context

No response

[github-actions]

Pinging code owners:

• receiver/prometheus: @Aneurysm9 @dashpole

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

[dashpole]

We are going to need the collector configuration to be able to help.