metrics: fix steal_count docs, add steal_operations (#5330)

tokio/src/runtime/metrics/batch.rs

@@ -11,9 +11,12 @@ pub(crate) struct MetricsBatch {
     /// Number of times the worker woke w/o doing work.
     noop_count: u64,
 
-    /// Number of times stolen.
+    /// Number of tasks stolen.
     steal_count: u64,
 
+    /// Number of times tasks where stolen.
+    steal_operations: u64,
+
     /// Number of tasks that were polled by the worker.
     poll_count: u64,
 
@@ -39,6 +42,7 @@ impl MetricsBatch {
             park_count: 0,
             noop_count: 0,
             steal_count: 0,
+            steal_operations: 0,
             poll_count: 0,
             poll_count_on_last_park: 0,
             local_schedule_count: 0,
@@ -52,6 +56,9 @@ impl MetricsBatch {
         worker.park_count.store(self.park_count, Relaxed);
         worker.noop_count.store(self.noop_count, Relaxed);
         worker.steal_count.store(self.steal_count, Relaxed);
+        worker
+            .steal_operations
+            .store(self.steal_operations, Relaxed);
         worker.poll_count.store(self.poll_count, Relaxed);
 
         worker
@@ -98,6 +105,10 @@ cfg_rt_multi_thread! {
             self.steal_count += by as u64;
         }
 
+        pub(crate) fn incr_steal_operations(&mut self) {
+            self.steal_operations += 1;
+        }
+
         pub(crate) fn incr_overflow_count(&mut self) {
             self.overflow_count += 1;
         }

tokio/src/runtime/metrics/mock.rs

@@ -38,6 +38,7 @@ impl MetricsBatch {
 cfg_rt_multi_thread! {
     impl MetricsBatch {
         pub(crate) fn incr_steal_count(&mut self, _by: u16) {}
+        pub(crate) fn incr_steal_operations(&mut self) {}
         pub(crate) fn incr_overflow_count(&mut self) {}
     }
 }

tokio/src/runtime/metrics/runtime.rs

@@ -210,10 +210,57 @@ impl RuntimeMetrics {
             .load(Relaxed)
     }
 
+    /// Returns the number of tasks the given worker thread stole from
+    /// another worker thread.
+    ///
+    /// This metric only applies to the **multi-threaded** runtime and will
+    /// always return `0` when using the current thread runtime.
+    ///
+    /// The worker steal count starts at zero when the runtime is created and
+    /// increases by `N` each time the worker has processed its scheduled queue
+    /// and successfully steals `N` more pending tasks from another worker.
+    ///
+    /// The counter is monotonically increasing. It is never decremented or
+    /// reset to zero.
+    ///
+    /// # Arguments
+    ///
+    /// `worker` is the index of the worker being queried. The given value must
+    /// be between 0 and `num_workers()`. The index uniquely identifies a single
+    /// worker and will continue to identify the worker throughout the lifetime
+    /// of the runtime instance.
+    ///
+    /// # Panics
+    ///
+    /// The method panics when `worker` represents an invalid worker, i.e. is
+    /// greater than or equal to `num_workers()`.
+    ///
+    /// # Examples
+    ///
+    /// ```
+    /// use tokio::runtime::Handle;
+    ///
+    /// #[tokio::main]
+    /// async fn main() {
+    ///     let metrics = Handle::current().metrics();
+    ///
+    ///     let n = metrics.worker_steal_count(0);
+    ///     println!("worker 0 has stolen {} tasks", n);
+    /// }
+    /// ```
+    pub fn worker_steal_count(&self, worker: usize) -> u64 {
+        self.handle
+            .inner
+            .worker_metrics(worker)
+            .steal_count
+            .load(Relaxed)
+    }
+
     /// Returns the number of times the given worker thread stole tasks from
     /// another worker thread.
     ///
-    /// This metric only applies to the **multi-threaded** runtime and will always return `0` when using the current thread runtime.
+    /// This metric only applies to the **multi-threaded** runtime and will
+    /// always return `0` when using the current thread runtime.
     ///
     /// The worker steal count starts at zero when the runtime is created and
     /// increases by one each time the worker has processed its scheduled queue
@@ -243,15 +290,15 @@ impl RuntimeMetrics {
     /// async fn main() {
     ///     let metrics = Handle::current().metrics();
     ///
-    ///     let n = metrics.worker_noop_count(0);
+    ///     let n = metrics.worker_steal_operations(0);
     ///     println!("worker 0 has stolen tasks {} times", n);
     /// }
     /// ```
-    pub fn worker_steal_count(&self, worker: usize) -> u64 {
+    pub fn worker_steal_operations(&self, worker: usize) -> u64 {
         self.handle
             .inner
             .worker_metrics(worker)
-            .steal_count
+            .steal_operations
             .load(Relaxed)
     }
 
@@ -328,8 +375,8 @@ impl RuntimeMetrics {
     /// async fn main() {
     ///     let metrics = Handle::current().metrics();
     ///
-    ///     let n = metrics.worker_poll_count(0);
-    ///     println!("worker 0 has polled {} tasks", n);
+    ///     let n = metrics.worker_total_busy_duration(0);
+    ///     println!("worker 0 was busy for a total of {:?}", n);
     /// }
     /// ```
     pub fn worker_total_busy_duration(&self, worker: usize) -> Duration {

tokio/src/runtime/metrics/worker.rs

@@ -17,9 +17,12 @@ pub(crate) struct WorkerMetrics {
     /// Number of times the worker woke then parked again without doing work.
     pub(crate) noop_count: AtomicU64,
 
-    /// Number of times the worker attempted to steal.
+    /// Number of tasks the worker stole.
     pub(crate) steal_count: AtomicU64,
 
+    /// Number of times the worker stole
+    pub(crate) steal_operations: AtomicU64,
+
     /// Number of tasks the worker polled.
     pub(crate) poll_count: AtomicU64,
 
@@ -43,6 +46,7 @@ impl WorkerMetrics {
             park_count: AtomicU64::new(0),
             noop_count: AtomicU64::new(0),
             steal_count: AtomicU64::new(0),
+            steal_operations: AtomicU64::new(0),
             poll_count: AtomicU64::new(0),
             overflow_count: AtomicU64::new(0),
             busy_duration_total: AtomicU64::new(0),

tokio/src/runtime/scheduler/multi_thread/queue.rs

@@ -353,6 +353,7 @@ impl<T> Steal<T> {
         }
 
         dst_metrics.incr_steal_count(n as u16);
+        dst_metrics.incr_steal_operations();
 
         // We are returning a task here
         n -= 1;