Skip to content

[AMORO-4089] Fix optimizing process stuck permanently after AMS restart #4090

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
j1wonpark wants to merge 1 commit into
base: master
Choose a base branch
from
+34 −10
Open

[AMORO-4089] Fix optimizing process stuck permanently after AMS restart #4090

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 5 additions & 4 deletions amoro-ams/src/main/java/org/apache/amoro/server/DefaultOptimizingService.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -490,16 +490,17 @@ public void run() {
OptimizerKeepingTask keepingTask = suspendingQueue.take();
String token = keepingTask.getToken();
boolean isExpired = !keepingTask.tryKeeping();
if (isExpired) {
LOG.info("Optimizer {} has been expired, unregister it", keepingTask.getOptimizer());
unregisterOptimizer(token);
}
Optional.ofNullable(keepingTask.getQueue())
.ifPresent(
queue ->
queue
.collectTasks(buildSuspendingPredication(authOptimizers.keySet()))
.forEach(task -> retryTask(task, queue)));
if (isExpired) {
LOG.info("Optimizer {} has been expired, unregister it", keepingTask.getOptimizer());
unregisterOptimizer(token);
} else {
if (!isExpired) {
LOG.debug("Optimizer {} is being touched, keep it", keepingTask.getOptimizer());
keepInTouch(keepingTask.getOptimizer());
}
Expand Down
8 changes: 8 additions & 0 deletions amoro-ams/src/main/java/org/apache/amoro/server/optimizing/OptimizingQueue.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -912,6 +912,14 @@ private void loadTaskRuntimes(OptimizingProcess optimizingProcess) {
taskMap.put(taskRuntime.getTaskId(), taskRuntime);
if (taskRuntime.getStatus() == TaskRuntime.Status.PLANNED) {
taskQueue.offer(taskRuntime);
} else if (taskRuntime.getStatus() == TaskRuntime.Status.SCHEDULED
|| taskRuntime.getStatus() == TaskRuntime.Status.ACKED) {
LOG.info(
"Reset task {} from {} to PLANNED during recovery",
taskRuntime.getTaskId(),
taskRuntime.getStatus());
taskRuntime.reset();
taskQueue.offer(taskRuntime);
} else if (taskRuntime.getStatus() == TaskRuntime.Status.FAILED) {
retryTask(taskRuntime);
}
Expand Down
27 changes: 21 additions & 6 deletions amoro-ams/src/test/java/org/apache/amoro/server/TestDefaultOptimizingService.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -248,7 +248,9 @@ public void testTouchTimeout() throws InterruptedException {
Assertions.assertThrows(PluginRetryAuthException.class, () -> optimizingService().touch(token));
Assertions.assertThrows(
PluginRetryAuthException.class, () -> optimizingService().pollTask(token, THREAD_ID));
assertTaskStatus(TaskRuntime.Status.SCHEDULED);
// After optimizer expires, its tasks are immediately reset to PLANNED
// because unregister happens before task scan in OptimizerKeeper
assertTaskStatus(TaskRuntime.Status.PLANNED);
token = optimizingService().authenticate(buildRegisterInfo());
toucher = new Toucher();
Thread.sleep(1000);
Expand Down Expand Up @@ -333,13 +335,19 @@ public void testReloadScheduledTask() {
OptimizingTask task = optimizingService().pollTask(token, THREAD_ID);
Assertions.assertNotNull(task);

// After reload, SCHEDULED tasks are reset to PLANNED during recovery
reload();
assertTaskStatus(TaskRuntime.Status.SCHEDULED);
optimizingService().ackTask(token, THREAD_ID, task.getTaskId());
assertTaskStatus(TaskRuntime.Status.PLANNED);

// Re-poll the task to get it scheduled again
OptimizingTask task2 = optimizingService().pollTask(token, THREAD_ID);
Assertions.assertNotNull(task2);
Assertions.assertEquals(task2.getTaskId(), task.getTaskId());
optimizingService().ackTask(token, THREAD_ID, task2.getTaskId());

TaskRuntime taskRuntime =
optimizingService().listTasks(defaultResourceGroup().getName()).get(0);
optimizingService().completeTask(token, buildOptimizingTaskResult(task.getTaskId()));
optimizingService().completeTask(token, buildOptimizingTaskResult(task2.getTaskId()));
assertTaskCompleted(taskRuntime);
}

Expand All @@ -350,12 +358,19 @@ public void testReloadAckTask() {
Assertions.assertNotNull(task);
optimizingService().ackTask(token, THREAD_ID, task.getTaskId());

// After reload, ACKED tasks are reset to PLANNED during recovery
reload();
assertTaskStatus(TaskRuntime.Status.ACKED);
assertTaskStatus(TaskRuntime.Status.PLANNED);

// Re-poll and re-ack the task
OptimizingTask task2 = optimizingService().pollTask(token, THREAD_ID);
Assertions.assertNotNull(task2);
Assertions.assertEquals(task2.getTaskId(), task.getTaskId());
optimizingService().ackTask(token, THREAD_ID, task2.getTaskId());

TaskRuntime<?> taskRuntime =
optimizingService().listTasks(defaultResourceGroup().getName()).get(0);
optimizingService().completeTask(token, buildOptimizingTaskResult(task.getTaskId()));
optimizingService().completeTask(token, buildOptimizingTaskResult(task2.getTaskId()));
assertTaskCompleted(taskRuntime);
}

Expand Down
Loading