feat(device/cpu): aggregate multi-socket zones into single zone

[sthaha]

This commit implements AggregatedZone to consolidate multiple EnergyZones with same name (e.g., package zones in multi-socket systems) into single zone. The aggregation also handles counter wrapping and is transparent to the caller. This enables power attribution across multi-socket systems while maintaining compatibility with single-socket deployments.

Additionally it also solves the zone label which used to have the index suffix as a hack.

Key changes:

• New AggregatedZone type that sums energy from multiple zones of the same type
• Proper handling of counter wrapping at MaxEnergy
• Thread-safe energy aggregation with overflow protection

[codecov]

Codecov Report

Attention: Patch coverage is 96.36364% with 4 lines in your changes missing coverage. Please review.

Project coverage is 92.39%. Comparing base (2c428ac) to head (bdce7f3).
Report is 4 commits behind head on reboot.

Files with missing lines	Patch %	Lines
internal/device/energy_zone.go	94.73%	3 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           reboot    #2183      +/-   ##
==========================================
- Coverage   92.41%   92.39%   -0.02%     
==========================================
  Files          37       38       +1     
  Lines        3809     3907      +98     
==========================================
+ Hits         3520     3610      +90     
- Misses        230      236       +6     
- Partials       59       61       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[vprashar2929]

Tested on multi-socket machine.

Available RAPL domains on machine:

root@aa4a13c3:/home/ghrunner/kepler/compose/dev# for d in /sys/devices/virtual/powercap/intel-rapl/intel-rapl:*; do   [ -f "$d/name" ] && echo "${d#/sys/devices/virtual/powercap/intel-rapl/}: $(cat "$d/name")";   for s in "$d"/intel-rapl:*; do     [ -f "$s/name" ] && echo "${s#/sys/devices/virtual/powercap/intel-rapl/}: $(cat "$s/name")";   done; done
intel-rapl:0: package-0
intel-rapl:0/intel-rapl:0:0: dram
intel-rapl:1: package-1
intel-rapl:1/intel-rapl:1:0: dram

Kepler creating aggregated zone:

kepler-dev-1  | time=2025-07-01T12:47:56.812Z level=DEBUG source=internal/device/rapl_sysfs_power_meter.go:177 msg="Created aggregated zone" service=rapl name=package zone_count=2 zones="[package-0 package-1]"
kepler-dev-1  | time=2025-07-01T12:47:56.812Z level=DEBUG source=internal/device/rapl_sysfs_power_meter.go:177 msg="Created aggregated zone" service=rapl name=dram zone_count=2 zones="[dram-0 dram-1]"

Comparing against node exporter

[vprashar2929]

@sthaha Right now, if I want to filter only the package-0 zone, it's not possible?

[sthaha]

@sthaha Right now, if I want to filter only the package-0 zone, it's not possible?

Thats right ... but if you run a version before this you should be able to see the values.

[vimalk78]

is there an option to not aggregate in some rare case?

[sthaha]

Good question and I think we will need that when/if kepler supports measuring individual socket consumption and is able to attribute that to process running on those sockets.

Or are you talking about some other requirement that we may have at the moment?

[vimalk78]

some reason we cant think of now. the default can be to aggregate, but if needed a config change (unexposed perhaps) will stop the aggregation.

[vimalk78]

why need mutex? sysfsRaplZone didn't need this