Struct CpuMetrics 

pub struct CpuMetrics {

    pub utilization_pct: f64,
    pub per_core_pct: Vec<f64>,
    pub utime_secs: f64,
    pub stime_secs: f64,
    pub process_count: u32,
    pub process_cores_used: Option<f64>,
    pub process_child_count: Option<u32>,
    pub process_utime_secs: Option<f64>,
    pub process_stime_secs: Option<f64>,
    pub process_rss_mib: Option<u64>,
    pub process_disk_read_bytes: Option<u64>,
    pub process_disk_write_bytes: Option<u64>,
    pub process_gpu_usage: Option<f64>,
    pub process_gpu_vram_mib: Option<f64>,
    pub process_gpu_utilized: Option<u32>,
    pub process_tree_pids: Vec<i32>,
}

CPU metrics derived from /proc/stat tick deltas.

Fields

utilization_pct: f64

Aggregate CPU utilization expressed as fractional cores in use (0.0..N_cores). e.g. 4.6 on a 16-core host means ~4.6 vCPUs are fully utilized. Not clamped; values very slightly above N_cores are valid under kernel rounding. N_cores is available via host discovery (host_vcpus).

per_core_pct: Vec<f64>

Per-core utilization indexed by logical CPU number (0.0–100.0 each).

utime_secs: f64

User+nice mode CPU time consumed across all cores in this interval (seconds). Equivalent to Δ(user+nice ticks) / ticks_per_second. Matches Python resource-tracker’s utime column.

stime_secs: f64

System mode CPU time consumed across all cores in this interval (seconds). Equivalent to Δ(system ticks) / ticks_per_second. Matches Python resource-tracker’s stime column.

process_count: u32

Number of processes currently in a runnable state (from /proc/stat procs_running). Matches Python resource-tracker’s processes column.

process_cores_used: Option<f64>

Fractional cores actively consumed by the tracked process tree (root process + all descendants), derived from /proc/<pid>/stat tick deltas divided by elapsed wall-clock ticks. e.g. 2.0 means the tree is consuming the equivalent of 2 full cores. None when no process PID is being tracked.

process_child_count: Option<u32>

Number of live descendant processes under the tracked root PID. Does not include the root process itself. None when no process PID is being tracked.

process_utime_secs: Option<f64>

User-mode CPU seconds consumed by the process tree this interval. Sum of utime tick deltas / ticks_per_second across all tree members. None when no PID is tracked.

process_stime_secs: Option<f64>

System-mode CPU seconds consumed by the process tree this interval. Sum of stime tick deltas / ticks_per_second across all tree members. None when no PID is tracked.

process_rss_mib: Option<u64>

Resident set size of the process tree (sum of VmRSS from /proc/pid/status) in MiB, sampled at each interval (not a delta). None when no PID is tracked.

process_disk_read_bytes: Option<u64>

Disk bytes actually read from storage by the process tree this interval. Delta of /proc/pid/io read_bytes across all tree members. None when no PID is tracked or /proc/pid/io is unreadable.

process_disk_write_bytes: Option<u64>

Disk bytes actually written to storage by the process tree this interval. Delta of /proc/pid/io write_bytes across all tree members. None when no PID is tracked or /proc/pid/io is unreadable.

process_gpu_usage: Option<f64>

Fractional GPUs actively consumed by the tracked process tree, expressed as the equivalent number of fully-utilized GPUs (same convention as process_cores_used). e.g. 0.5 means the tree is using half of one GPU. NVIDIA: derived from SM utilization via nvmlDeviceGetProcessUtilization (no accounting mode required), summed across matched PIDs and devices, then divided by 100. AMD: derived from drm-engine-gfx cumulative ns via FdInfoStat delta tracking. None when no GPU is present, NVML/AMD data is unavailable, or no samples returned.

process_gpu_vram_mib: Option<f64>

Total VRAM consumed by the tracked process tree across all GPUs (MiB). NVIDIA: sum of used_gpu_memory from NVML running-process lists. AMD: sum of drm-memory-vram from /proc/pid/fdinfo for matched devices. None when no PID is tracked or no GPU is present on the host.

process_gpu_utilized: Option<u32>

Number of GPUs on which at least one process in the tracked tree has allocated VRAM or appears in the running-process list. None when no PID is tracked or no GPU is present on the host.

process_tree_pids: Vec<i32>

PIDs in the tracked process tree (root + all descendants). Populated by CpuCollector; used by main.rs to query per-process GPU stats. Skipped in JSON/CSV output – internal routing field only.

Trait Implementations

impl Clone for CpuMetrics

fn clone(&self) -> CpuMetrics

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CpuMetrics

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for CpuMetrics

fn default() -> CpuMetrics

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for CpuMetrics

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for CpuMetrics

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.