fix(obs): isolate process sampler windows (#4492)

* fix(obs): isolate process sampler windows

Refs rustfs/backlog#1004
Refs rustfs/backlog#986

- add a reusable ProcessSampler so callers can own independent sysinfo refresh windows
- wire separate sampler instances for obs metrics scheduling and memory observability
- keep compatibility helpers while avoiding cross-task CPU and disk delta interference

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): import process sampler bundle helper

Refs rustfs/backlog#1004
Refs rustfs/backlog#986

- import collect_process_metric_bundle_with in the metrics scheduler
- drop the stale collect_process_metric_bundle import after switching scheduler sampling to independent process samplers

Co-Authored-By: heihutu <heihutu@gmail.com>

* fix(obs): move process sampler into blocking task

Refs rustfs/backlog#1004
Refs rustfs/backlog#986

- move the memory observability process sampler into the spawn_blocking closure
- satisfy the closure static lifetime required by tokio while keeping the isolated sampler design intact

Co-Authored-By: heihutu <heihutu@gmail.com>

---------

Co-authored-by: heihutu <heihutu@gmail.com>
This commit is contained in:
houseme
2026-07-08 23:51:34 +08:00
committed by GitHub
parent 6bdfdd164a
commit 6cb47049e8
7 changed files with 98 additions and 27 deletions
+51 -12
View File
@@ -15,21 +15,43 @@
use std::sync::{Mutex, OnceLock};
use sysinfo::{Pid, ProcessRefreshKind, ProcessStatus, ProcessesToUpdate, System};
static PROCESS_SYSTEM: OnceLock<Mutex<System>> = OnceLock::new();
static PROCESS_SAMPLER: OnceLock<Mutex<ProcessSampler>> = OnceLock::new();
#[inline]
fn current_pid() -> Pid {
Pid::from_u32(std::process::id())
}
#[inline]
fn process_system() -> &'static Mutex<System> {
PROCESS_SYSTEM.get_or_init(|| {
#[derive(Debug)]
pub struct ProcessSampler {
pid: Pid,
sys: System,
}
impl Default for ProcessSampler {
fn default() -> Self {
Self::new()
}
}
impl ProcessSampler {
#[inline]
pub fn new() -> Self {
let pid = current_pid();
let mut system = System::new();
system.refresh_processes_specifics(ProcessesToUpdate::Some(&[pid]), true, ProcessRefreshKind::everything());
Mutex::new(system)
})
let mut sys = System::new();
sys.refresh_processes_specifics(ProcessesToUpdate::Some(&[pid]), true, ProcessRefreshKind::everything());
Self { pid, sys }
}
#[inline]
pub fn snapshot_resource_and_system(&mut self) -> (ProcessResourceSnapshot, ProcessSystemSnapshot) {
snapshot_process_resource_and_system_with(self)
}
}
#[inline]
fn process_sampler() -> &'static Mutex<ProcessSampler> {
PROCESS_SAMPLER.get_or_init(|| Mutex::new(ProcessSampler::new()))
}
#[derive(Debug, Clone, Copy, Default, PartialEq)]
@@ -99,13 +121,21 @@ pub fn snapshot_process_system() -> ProcessSystemSnapshot {
/// Collect both resource and system snapshots in one sysinfo refresh.
#[inline]
pub fn snapshot_process_resource_and_system() -> (ProcessResourceSnapshot, ProcessSystemSnapshot) {
let mut sampler = process_sampler().lock().unwrap_or_else(|poisoned| poisoned.into_inner());
snapshot_process_resource_and_system_with(&mut sampler)
}
#[inline]
pub fn snapshot_process_resource_and_system_with(
sampler: &mut ProcessSampler,
) -> (ProcessResourceSnapshot, ProcessSystemSnapshot) {
let platform_stats = crate::snapshot_process_platform_stats();
let lock_snapshot = crate::snapshot_process_lock_counts();
let pid = current_pid();
let mut sys = process_system().lock().unwrap_or_else(|poisoned| poisoned.into_inner());
sys.refresh_processes_specifics(ProcessesToUpdate::Some(&[pid]), true, ProcessRefreshKind::everything());
sampler
.sys
.refresh_processes_specifics(ProcessesToUpdate::Some(&[sampler.pid]), true, ProcessRefreshKind::everything());
if let Some(process) = sys.process(pid) {
if let Some(process) = sampler.sys.process(sampler.pid) {
let disk_usage = process.disk_usage();
let status = ProcessStatusSnapshot::from(process.status());
let uptime_seconds = process.run_time();
@@ -163,4 +193,13 @@ mod tests {
let _ = snapshot_process_system();
let _ = snapshot_process_resource_and_system();
}
#[test]
fn independent_samplers_are_collectable() {
let mut sampler_a = ProcessSampler::new();
let mut sampler_b = ProcessSampler::new();
let _ = snapshot_process_resource_and_system_with(&mut sampler_a);
let _ = snapshot_process_resource_and_system_with(&mut sampler_b);
}
}