[
https://issues.apache.org/jira/browse/HDFS-17284?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17800328#comment-17800328
]
ASF GitHub Bot commented on HDFS-17284:
---------------------------------------
tasanuma commented on code in PR #6348:
URL: https://github.com/apache/hadoop/pull/6348#discussion_r1436119492
##########
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestDatanodeManager.java:
##########
@@ -1126,4 +1128,66 @@ public MockDfsNetworkTopology(){
super();
}
}
+
+ @Test
+ public void testComputeReconstructedTaskNum() throws IOException {
+ verifyComputeReconstructedTaskNum(100, 100, 150, 250, 100);
+ verifyComputeReconstructedTaskNum(200, 100000, 200000, 300000, 400000);
+ verifyComputeReconstructedTaskNum(1000000, 100, 150, 250, 100);
+ verifyComputeReconstructedTaskNum(14000000, 200, 200, 400, 200);
+
+ }
+ public void verifyComputeReconstructedTaskNum(int xmitsInProgress, int
numReplicationBlocks,
+ int maxTransfers, int numECTasksToBeReplicated, int
numBlocksToBeErasureCoded)
+ throws IOException {
+ FSNamesystem fsn = Mockito.mock(FSNamesystem.class);
+ Mockito.when(fsn.hasWriteLock()).thenReturn(true);
+ Configuration conf = new Configuration();
+ conf.setInt(DFSConfigKeys.DFS_NAMENODE_REPLICATION_MAX_STREAMS_KEY,
maxTransfers);
+ DatanodeManager dm = Mockito.spy(mockDatanodeManager(fsn, conf));
+
+ DatanodeDescriptor nodeInfo = Mockito.mock(DatanodeDescriptor.class);
+ Mockito.when(nodeInfo.isRegistered()).thenReturn(true);
+ Mockito.when(nodeInfo.getStorageInfos()).thenReturn(new
DatanodeStorageInfo[0]);
+
+
Mockito.when(nodeInfo.getNumberOfReplicateBlocks()).thenReturn(numReplicationBlocks);
+
Mockito.when(nodeInfo.getNumberOfECBlocksToBeReplicated()).thenReturn(numECTasksToBeReplicated);
+ Mockito.when(nodeInfo.getNumberOfBlocksToBeErasureCoded())
+ .thenReturn(numBlocksToBeErasureCoded);
+
+ // Create an ArgumentCaptor to capture the counts for numReplicationTasks,
+ // numEcReplicatedTasks,numECReconstructedTasks.
+ ArgumentCaptor<Integer> captor = ArgumentCaptor.forClass(Integer.class);
+ Mockito.when(nodeInfo.getErasureCodeCommand(ArgumentMatchers.anyInt()))
+ .thenReturn(Collections.nCopies(0, null));
+ Mockito.when(nodeInfo.getReplicationCommand(ArgumentMatchers.anyInt()))
+ .thenReturn(Collections.nCopies(0, null));
+ Mockito.when(nodeInfo.getECReplicatedCommand(ArgumentMatchers.anyInt()))
+ .thenReturn(Collections.nCopies(0, null));
+
+ DatanodeRegistration nodeReg = Mockito.mock(DatanodeRegistration.class);
+ Mockito.when(dm.getDatanode(nodeReg)).thenReturn(nodeInfo);
+
+
+ dm.handleHeartbeat(nodeReg, new StorageReport[1], "bp-123", 0, 0,
+ 10, xmitsInProgress, 0, null, SlowPeerReports.EMPTY_REPORT,
+ SlowDiskReports.EMPTY_REPORT);
+
+ Mockito.verify(nodeInfo).getReplicationCommand(captor.capture());
+ int numReplicationTasks = captor.getValue();
+
+ Mockito.verify(nodeInfo).getECReplicatedCommand(captor.capture());
+ int numEcReplicatedTasks = captor.getValue();
+
+ Mockito.verify(nodeInfo).getErasureCodeCommand(captor.capture());
+ int numECReconstructedTasks = captor.getValue();
+
+ // Verify that when DN xmitsInProgress exceeds maxTransfers,
+ // the number of tasks should be <= 0.
+ if(xmitsInProgress >= maxTransfers){
+ assertTrue(numReplicationTasks <= 0);
+ assertTrue(numEcReplicatedTasks <= 0);
+ assertTrue(numECReconstructedTasks <= 0);
+ }
Review Comment:
To ensure that `verifyComputeReconstructedTaskNum(200, 100000, 200000,
300000, 400000)` doesn't cause an overflow, it would be necessary to check the
reverse situation as well.
```suggestion
} else {
assertTrue(numReplicationTasks >= 0);
assertTrue(numEcReplicatedTasks >= 0);
assertTrue(numECReconstructedTasks >= 0);
}
```
Other than that, it looks good to me.
> Fix int overflow in calculating numEcReplicatedTasks and numReplicationTasks
> during block recovery
> --------------------------------------------------------------------------------------------------
>
> Key: HDFS-17284
> URL: https://issues.apache.org/jira/browse/HDFS-17284
> Project: Hadoop HDFS
> Issue Type: Bug
> Components: namenode
> Reporter: Hualong Zhang
> Assignee: Hualong Zhang
> Priority: Major
> Labels: pull-request-available
>
> Fix int overflow in calculating numEcReplicatedTasks and numReplicationTasks
> during block recovery
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: hdfs-issues-unsubscr...@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-h...@hadoop.apache.org
