I want to filter out (either including or excluding) hostmetrics’s CPU based on metric’s labels but this does not work well when I use the include/exclude syntax for filterprocessor.

Here are my config.yml:

receivers:
  hostmetrics:
    collection_interval: 10s
    scrapers:
      cpu:
        metrics:
          system.cpu.time:
            enabled: true

processors:
  filter/metric_attr:
    error_mode: ignore
    metrics:
      include:
        match_type: expr
        expressions:
          - Label("cpu") matches "^cpu0$"

exporters:
  prometheus:
    endpoint: localhost:8889
    namespace: otel-host-metrics

service:
  telemetry:
    logs:
      level: debug
  pipelines:
    metrics:
      receivers: [hostmetrics]
      processors: [filter/metric_attr]
      exporters: [prometheus]

What actually works is using OTTL syntax, but this can only drop metrics, it does not have the flexibility to include/exclude:

# This drops the metrics which has cpu attribute matches cpu1
  filter/drop_label:
    error_mode: ignore
    metrics:
      datapoint:
        - attributes["cpu"] == "cpu1"

For additional info, this is what hostmetrics CPU looks like when printed out using debug exporters with verbosity: detailed:

Metric #0
Descriptor:
     -> Name: system.cpu.time
     -> Description: Total seconds each logical CPU spent on each mode.
     -> Unit: s
     -> DataType: Sum
     -> IsMonotonic: true
     -> AggregationTemporality: Cumulative
NumberDataPoints #0
Data point attributes:
     -> cpu: Str(cpu0)
     -> state: Str(user)
StartTimestamp: 2024-07-13 17:45:32 +0000 UTC
Timestamp: 2024-07-15 07:01:21.909324181 +0000 UTC
Value: 364.330000