Kepler energy metrics are zeroes

[Lai-Kenny]

Hello, everyone.
My Environment:

Ubuntu	Linux Kernel	Kubernetes
20.04.6 LTS	5.4.0-148-generic x86_64	v1.16.0

I install the Kepler by the kepler-helm-chart. I try some version of kepler chart version(0.4.2 and 0.4.1)
However, I found some metrics(e.g. kepler_container_cgroupfs_cpu_usage_us_total, kepler_container_core_joules_total and so on) is alway 0. It would affect estimate energy consumed, right??
And I try to follow the Trouble Shooting of Kepler to debug it. But I check my version of cGroup is v2.

So, how should I debug it?

[yellowhat]

Hi,
does not seem like a helm chart issue.
@rootfs any suggestion?

Just a comment, from here seems that cgroup v2 are after ubuntu 21.10.

[Lai-Kenny]

Hi, @yellowhat @rootfs
I have updated my Environment
kepler chart version: 0.5.0
Ubuntu: 22.04.3 LTS
Linux Kernel: 5.15.0-78-generic x86_64
kubelet: 1.21.0
cgroup: v2
But my some metrics(e.g. kepler_container_cgroupfs_cpu_usage_us_total, kepler_container_core_joules_total and so on) is alway 0. The following is my kepler log

I0810 13:36:11.444138       1 gpu.go:46] Failed to init nvml, err: could not init nvml: error opening libnvidia-ml.so.1: libnvidia-ml.so.1: cannot open shared object file: No such file or directory
I0810 13:36:11.456218       1 exporter.go:155] Kepler running on version: a7a6cb1
I0810 13:36:11.456299       1 config.go:258] using gCgroup ID in the BPF program: true
I0810 13:36:11.456394       1 config.go:260] kernel version: 5.15
I0810 13:36:11.456441       1 exporter.go:179] EnabledBPFBatchDelete: true
I0810 13:36:11.456485       1 rapl_msr_util.go:129] failed to open path /dev/cpu/0/msr: no such file or directory
I0810 13:36:11.456556       1 power.go:64] Not able to obtain power, use estimate method
I0810 13:36:11.456585       1 redfish.go:169] failed to get redfish credential file path
I0810 13:36:11.456601       1 power.go:55] use acpi to obtain power
I0810 13:36:11.456790       1 acpi.go:67] Could not find any ACPI power meter path. Is it a VM?
I0810 13:36:11.467127       1 exporter.go:198] Initializing the GPU collector
I0810 13:36:17.469943       1 watcher.go:66] Using in cluster k8s config
I0810 13:36:17.571852       1 bpf_perf.go:123] LibbpfBuilt: false, BccBuilt: true
cannot attach kprobe, probe entry may not exist
I0810 13:36:18.414147       1 bcc_attacher.go:186] Successfully load eBPF module from bcc with option: [-DMAP_SIZE=10240 -DNUM_CPUS=4 -DSET_GROUP_ID]
I0810 13:36:18.440766       1 exporter.go:251] Started Kepler in 6.984594733s

[rootfs]

@Lai-Kenny can you check if cpu accounting is turned on? please follow the threads here