Malicious log flow status caused by buffer metrics

[a-thaler]

Description The metrics for the log buffer are based on a custom sidecar tracking the underlying filesystem. That sidecar is not correlating the buffer instance with an actual logpipeline. With that, the flow health status of LogPipeline A can get negative if the buffer of LogPipeline B is running full. Also, a leftover buffer of a previous pipeline can stimulate this.

Criterias

• The flow status of a pipeline should be isolated on the pipeline itself by looking only at the metrics of the dedicated buffer

[github-actions[bot]]

This issue has been automatically marked as stale due to the lack of recent activity. It will soon be closed if no further activity occurs. Thank you for your contributions.