[GitHub] apucher closed pull request #3634: [TE] task - randomize execution order to increase parallel throughput

[GitBox <gi...@apache.org>]

apucher closed pull request #3634: [TE] task - randomize execution order to increase parallel throughput
URL: https://github.com/apache/incubator-pinot/pull/3634
 
 
   

This is a PR merged from a forked repository.
As GitHub hides the original diff on merge, it is displayed below for
the sake of provenance:

As this is a foreign pull request (from a fork), the diff is supplied
below (as it won't show otherwise due to GitHub magic):

diff --git a/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/anomaly/task/TaskDriver.java b/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/anomaly/task/TaskDriver.java
index 5b43998905..3e8104267a 100644
--- a/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/anomaly/task/TaskDriver.java
+++ b/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/anomaly/task/TaskDriver.java
@@ -21,6 +21,7 @@
 import com.linkedin.thirdeye.detector.email.filter.AlertFilterFactory;
 
 import java.util.ArrayList;
+import java.util.Collections;
 import java.util.HashSet;
 import java.util.List;
 import java.util.Random;
@@ -134,6 +135,7 @@ private TaskDTO acquireTask() {
       List<TaskDTO> anomalyTasks = new ArrayList<>();
       boolean hasFetchError = false;
       try {
+        // randomize fetching head and tail to reduce synchronized patterns across threads (and hosts)
         boolean orderAscending = System.currentTimeMillis() % 2 == 0;
         anomalyTasks = taskDAO
             .findByStatusOrderByCreateTime(TaskStatus.WAITING, driverConfiguration.getTaskFetchSizeCap(),
@@ -147,6 +149,9 @@ private TaskDTO acquireTask() {
       if (CollectionUtils.isNotEmpty(anomalyTasks)) {
         LOG.info("Thread {} : Found {} tasks in waiting state", Thread.currentThread().getId(), anomalyTasks.size());
 
+        // shuffle candidate tasks to avoid synchronized patterns across threads (and hosts)
+        Collections.shuffle(anomalyTasks);
+
         for (int i = 0; i < anomalyTasks.size() && !shutdown; i++) {
           TaskDTO anomalyTaskSpec = anomalyTasks.get(i);
 
diff --git a/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/datalayer/bao/jdbc/TaskManagerImpl.java b/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/datalayer/bao/jdbc/TaskManagerImpl.java
index b44db6d614..141a330410 100644
--- a/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/datalayer/bao/jdbc/TaskManagerImpl.java
+++ b/thirdeye/thirdeye-pinot/src/main/java/com/linkedin/thirdeye/datalayer/bao/jdbc/TaskManagerImpl.java
@@ -38,10 +38,10 @@
 public class TaskManagerImpl extends AbstractManagerImpl<TaskDTO> implements TaskManager {
 
   private static final String FIND_BY_STATUS_ORDER_BY_CREATE_TIME_ASC =
-      " WHERE status = :status order by startTime asc limit 5";
+      " WHERE status = :status order by startTime asc limit 10";
 
   private static final String FIND_BY_STATUS_ORDER_BY_CREATE_TIME_DESC =
-      " WHERE status = :status order by startTime desc limit 5";
+      " WHERE status = :status order by startTime desc limit 10";
 
   public TaskManagerImpl() {
     super(TaskDTO.class, TaskBean.class);


 

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@pinot.apache.org
For additional commands, e-mail: dev-help@pinot.apache.org