fix: container_fs metrics {device} has no label or info-metric to associate it with a volumeattachment or persistentvolumeclaim

[tiemma]

This pull request adds improved tracking and reporting of filesystem mountpoints and container paths for disk metrics, enabling more accurate association of storage devices with container volumes (such as PersistentVolumeClaims) in Prometheus metrics. The changes include new fields in filesystem stats, a new Prometheus metric, and logic to extract volume names from mount paths, along with expanded test coverage.

Filesystem mountpoint and container path enhancements:

• Added Mountpoint and ContainerPath fields to the FsStats struct in info/v1/container.go and updated the Fs struct in fs/types.go to track the host-side mountpoint for each filesystem device. [1] [2]
• Updated fs/fs.go and container/raw/handler.go to populate mountpoint and container path information when collecting filesystem stats, including mapping host mount directories to container paths using external mounts configuration. [1] [2] [3] [4]

Prometheus metrics improvements:

• Introduced a new Prometheus metric container_fs_device_info to expose device, mountpoint, container path, volume name, and volume type for container filesystems, allowing correlation with other disk metrics. [1] [2] [3] [4]

Volume name extraction logic:

• Added parseMountPointIntoVolumePath function in metrics/prometheus.go to extract volume names and types from kubelet volume mount paths, supporting CSI and NFS volumes.

Expanded test coverage:

• Added new unit tests for mountpoint handling and volume name extraction, including test cases for CSI, NFS, and non-standard paths in container/raw/handler_test.go and metrics/prometheus_test.go. [1] [2] [3]

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[tiemma]

Currently, there is no way to associate a container workload with a pvc or volume attachment, the goal is to provide volume information based on the mount point to allow the user join on these metrics from kube-state-metrics or other programmatic forms to simplify associations.

From the new prometheus metric container_fs_device_info, other users can associate volume to pvc associations if the device is a valid csi mount point.

Related issues:

• container_fs metrics {device} has no label or info-metric to associate it with a volumeattachment or persistentvolumeclaim #3588
• Add I/O metric for persistentvolumeclaim and persistentvolume #3074
• Support diskIO stats for persistent volumes #1445
• Unable to see container's metrics about external volumes attached (k8s persistent volumes) #1702
• when computer container cpu usage through the metrics/cadvisor, the value larger than limited value and top command's value #2362