clebertsuconic commented on a change in pull request #2741: ARTEMIS-2415
JDBCJournal miss pending tasks during shutdown
URL: https://github.com/apache/activemq-artemis/pull/2741#discussion_r305345667
##########
File path:
artemis-server/src/test/java/org/apache/activemq/artemis/core/persistence/impl/journal/JournalStorageManagerTest.java
##########
@@ -212,4 +212,36 @@ public void
testAddBytesToLargeMessageNotLeakingByteBuffer() throws Exception {
}
}
+ @Test
+ public void testDeletingLargeMessagePendingTasksOnShutdown() throws
Exception {
+ if (journalType == JournalType.ASYNCIO) {
+ assumeTrue("AIO is not supported on this platform",
AIOSequentialFileFactory.isSupported());
+ }
+ final Configuration configuration =
createDefaultInVMConfig().setJournalType(journalType);
+ final ExecutorFactory executorFactory = new
OrderedExecutorFactory(executor);
+ final ExecutorFactory ioExecutorFactory = new
OrderedExecutorFactory(ioExecutor);
+ final JournalStorageManager manager = new
JournalStorageManager(configuration, null, executorFactory, null,
ioExecutorFactory);
+ manager.start();
+ manager.loadBindingJournal(new ArrayList<>(), new ArrayList<>(), new
ArrayList<>());
+ final PostOffice postOffice = mock(PostOffice.class);
+ final JournalLoader journalLoader = mock(JournalLoader.class);
+ manager.loadMessageJournal(postOffice, null, null, null, null, null,
null, journalLoader);
+ final LargeServerMessage largeMessage =
manager.createLargeMessage(manager.generateID() + 1, new
CoreMessage().setDurable(true));
+ final SequentialFile file = largeMessage.getFile();
+ manager.getContext(true).storeLineUp();
+ manager.deleteLargeMessageFile(largeMessage);
+ final CountDownLatch latch = new CountDownLatch(1);
+ Executors.newSingleThreadExecutor().execute(() -> {
+ try {
+ manager.stop();
+ latch.countDown();
+ } catch (Exception ignore) {
+ }
+ });
+ Thread.sleep(500);
Review comment:
Why you need a sleep? I have been beaten pretty badly by Thread.sleep() on
the testsuite many many times. Can't we avoid it with a latch, a Wait..
anything else?
Unless this is testing timeouts, I wouldn't use a wait.
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
With regards,
Apache Git Services
