cleanup

Author: mhl-b

benchmarks/src/main/java/org/elasticsearch/benchmark/common/util/concurrent/ThreadPoolUtilizationBenchmark.java

@@ -27,7 +27,7 @@
 
 import java.util.concurrent.TimeUnit;
 
-@Threads(12)
+@Threads(Threads.MAX)
 @Warmup(iterations = 3, time = 200, timeUnit = TimeUnit.MILLISECONDS)
 @Measurement(iterations = 5, time = 600, timeUnit = TimeUnit.MILLISECONDS)
 @BenchmarkMode(Mode.SampleTime)
@@ -36,13 +36,13 @@
 @Fork(1)
 public class ThreadPoolUtilizationBenchmark {
 
-    @Param({ "10000" })
+    @Param({ "1000" })
     private int callIntervalTicks;
 
     /**
      * This makes very little difference, all the overhead is in the synchronization
      */
-    @Param({ "10" })
+    @Param({ "30000" })
     private int utilizationIntervalMs;
     private TaskExecutionTimeTrackingEsThreadPoolExecutor.FramedTimeTracker timeTracker;
 

server/src/main/java/org/elasticsearch/common/util/concurrent/TaskExecutionTimeTrackingEsThreadPoolExecutor.java

@@ -157,7 +157,7 @@ public TaskTrackingConfig trackingConfig() {
      * Utilization is measured as {@code all-threads-total-execution-time / (total-thread-count * interval)}.
      * This metric is updated once per interval, and returns last completed measurement. For example:
      * if interval is 30 seconds, at clock time 00:30-01:00 it will return utilization from 00:00-00:30.
-     * Thou there is no synchronization with clocks and system time.
+     * There is no synchronization with clocks and system time.
      *
      * If caller needs longer intervals it should poll on every tracker-interval and aggregate on it's own. Another option is to extend
      * framedTimeTracker to remember multiple past frames, and return aggregated view from here.
@@ -254,7 +254,7 @@ public boolean trackingMaxQueueLatency() {
     }
 
     /**
-     * Tracks treads execution in continuous, non-overlapping, and even time frames. Provides accurate total execution time measurement
+     * Tracks threads execution in continuous, non-overlapping, and even time frames. Provides accurate total execution time measurement
      * for past frames, specifically previous frame (now - 1 frame) to measure utilization.
      *
      * Can be extended to remember multiple past frames.
@@ -327,7 +327,7 @@ public void startTask() {
             var now = timeNow.get();
             updateFrame0(now);
             ongoingTasks.incrementAndGet();
-            currentTime.updateAndGet((t) -> t + (currentFrame.get() + 1) * interval - now);
+            currentTime.addAndGet((now + 1) * interval - now);
             rwlock.readLock().unlock();
         }
 
@@ -339,7 +339,7 @@ public void endTask() {
             var now = timeNow.get();
             updateFrame0(now);
             ongoingTasks.decrementAndGet();
-            currentTime.updateAndGet((t) -> t - (currentFrame.get() + 1) * interval + now);
+            currentTime.addAndGet(-((now + 1) * interval - now));
             rwlock.readLock().unlock();
         }