[HDFS-9580] TestComputeInvalidateWork#testDatanodeReRegistration failed due to unexpected number of invalidate blocks.

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 3.0.0-alpha1
• Fix Version/s: 2.8.0, 3.0.0-alpha1
• Component/s: datanode, namenode, test

Environment

Jenkins

• Hadoop Flags: Reviewed
• Assignee: Wei-Chiu Chuang
• Reporter: Wei-Chiu Chuang
• Created: 19/Dec/15 14:47
• Updated: 02/Oct/19 17:14
• Resolved: 22/Dec/15 06:39

Description

The failure appeared in the trunk jenkins job.
https://builds.apache.org/job/Hadoop-Hdfs-trunk/2646/

Error Message

Expected invalidate blocks to be the number of DNs expected:<3> but was:<2>

Stacktrace

java.lang.AssertionError: Expected invalidate blocks to be the number of DNs expected:<3> but was:<2>
	at org.junit.Assert.fail(Assert.java:88)
	at org.junit.Assert.failNotEquals(Assert.java:743)
	at org.junit.Assert.assertEquals(Assert.java:118)
	at org.junit.Assert.assertEquals(Assert.java:555)
	at org.apache.hadoop.hdfs.server.blockmanagement.TestComputeInvalidateWork.testDatanodeReRegistration(TestComputeInvalidateWork.java:160)

I think there could be a race condition between creating a file and shutting down data nodes, which failed the test.

2015-12-19 07:11:02,765 [PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=LAST_IN_PIPELINE, downstreams=0:[]] INFO  datanode.DataNode (BlockReceiver.java:run(1404)) - PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=LAST_IN_PIPELINE, downstreams=0:[] terminating
2015-12-19 07:11:02,768 [PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE] INFO  DataNode.clienttrace (BlockReceiver.java:finalizeBlock(1431)) - src: /127.0.0.1:45655, dest: /127.0.0.1:54890, bytes: 134217728, op: HDFS_WRITE, cliID: DFSClient_NONMAPREDUCE_147911011_935, offset: 0, srvID: 6a13ec05-e1c1-4086-8a4d-d5a09636afcd, blockid: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, duration: 954174423
2015-12-19 07:11:02,768 [PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE] INFO  datanode.DataNode (BlockReceiver.java:run(1404)) - PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE terminating
2015-12-19 07:11:02,772 [PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE] INFO  DataNode.clienttrace (BlockReceiver.java:finalizeBlock(1431)) - src: /127.0.0.1:33252, dest: /127.0.0.1:54426, bytes: 134217728, op: HDFS_WRITE, cliID: DFSClient_NONMAPREDUCE_147911011_935, offset: 0, srvID: d81751db-02a9-48fe-b697-77623048784b, blockid: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, duration: 957463510
2015-12-19 07:11:02,772 [PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE] INFO  datanode.DataNode (BlockReceiver.java:run(1404)) - PacketResponder: BP-1551077294-67.195.81.149-1450509060247:blk_1073741825_1001, type=HAS_DOWNSTREAM_IN_PIPELINE terminating
2015-12-19 07:11:02,782 [IPC Server handler 4 on 36404] INFO  blockmanagement.BlockManager (BlockManager.java:checkBlocksProperlyReplicated(3871)) - BLOCK* blk_1073741825_1001 is not COMPLETE (ucState = COMMITTED, replication# = 0 <  minimum = 1) in file /testRR
2015-12-19 07:11:02,783 [IPC Server handler 4 on 36404] INFO  namenode.EditLogFileOutputStream (EditLogFileOutputStream.java:flushAndSync(200)) - Nothing to flush
2015-12-19 07:11:02,783 [IPC Server handler 4 on 36404] INFO  namenode.EditLogFileOutputStream (EditLogFileOutputStream.java:flushAndSync(200)) - Nothing to flush
2015-12-19 07:11:03,190 [IPC Server handler 8 on 36404] INFO  hdfs.StateChange (FSNamesystem.java:completeFile(2557)) - DIR* completeFile: /testRR is closed by DFSClient_NONMAPREDUCE_147911011_935

Issue Links

• is related to: HDFS-10816 TestComputeInvalidateWork#testDatanodeReRegistration fails due to race between test and replication monitor