/** Test to check that a DN goes down when all its volumes have failed. */
@Test
public void testShutdown() throws Exception {
if (System.getProperty("os.name").startsWith("Windows")) {
/**
* This test depends on OS not allowing file creations on a directory that does not have write
* permissions for the user. Apparently it is not the case on Windows (at least under Cygwin),
* and possibly AIX. This is disabled on Windows.
*/
return;
}
// Bring up two more datanodes
cluster.startDataNodes(conf, 2, true, null, null);
cluster.waitActive();
final int dnIndex = 0;
String bpid = cluster.getNamesystem().getBlockPoolId();
File storageDir = cluster.getInstanceStorageDir(dnIndex, 0);
File dir1 = MiniDFSCluster.getRbwDir(storageDir, bpid);
storageDir = cluster.getInstanceStorageDir(dnIndex, 1);
File dir2 = MiniDFSCluster.getRbwDir(storageDir, bpid);
try {
// make the data directory of the first datanode to be readonly
assertTrue("Couldn't chmod local vol", dir1.setReadOnly());
assertTrue("Couldn't chmod local vol", dir2.setReadOnly());
// create files and make sure that first datanode will be down
DataNode dn = cluster.getDataNodes().get(dnIndex);
for (int i = 0; dn.isDatanodeUp(); i++) {
Path fileName = new Path("/test.txt" + i);
DFSTestUtil.createFile(fs, fileName, 1024, (short) 2, 1L);
DFSTestUtil.waitReplication(fs, fileName, (short) 2);
fs.delete(fileName, true);
}
} finally {
// restore its old permission
FileUtil.setWritable(dir1, true);
FileUtil.setWritable(dir2, true);
}
}
/**
* Test that when there is a failure replicating a block the temporary and meta files are cleaned
* up and subsequent replication succeeds.
*/
@Test
public void testReplicationError() throws Exception {
// create a file of replication factor of 1
final Path fileName = new Path("/test.txt");
final int fileLen = 1;
DFSTestUtil.createFile(fs, fileName, 1, (short) 1, 1L);
DFSTestUtil.waitReplication(fs, fileName, (short) 1);
// get the block belonged to the created file
LocatedBlocks blocks =
NameNodeAdapter.getBlockLocations(
cluster.getNameNode(), fileName.toString(), 0, (long) fileLen);
assertEquals("Should only find 1 block", blocks.locatedBlockCount(), 1);
LocatedBlock block = blocks.get(0);
// bring up a second datanode
cluster.startDataNodes(conf, 1, true, null, null);
cluster.waitActive();
final int sndNode = 1;
DataNode datanode = cluster.getDataNodes().get(sndNode);
// replicate the block to the second datanode
InetSocketAddress target = datanode.getXferAddress();
Socket s = new Socket(target.getAddress(), target.getPort());
// write the header.
DataOutputStream out = new DataOutputStream(s.getOutputStream());
DataChecksum checksum = DataChecksum.newDataChecksum(DataChecksum.Type.CRC32, 512);
new Sender(out)
.writeBlock(
block.getBlock(),
StorageType.DEFAULT,
BlockTokenSecretManager.DUMMY_TOKEN,
"",
new DatanodeInfo[0],
new StorageType[0],
null,
BlockConstructionStage.PIPELINE_SETUP_CREATE,
1,
0L,
0L,
0L,
checksum,
CachingStrategy.newDefaultStrategy(),
false);
out.flush();
// close the connection before sending the content of the block
out.close();
// the temporary block & meta files should be deleted
String bpid = cluster.getNamesystem().getBlockPoolId();
File storageDir = cluster.getInstanceStorageDir(sndNode, 0);
File dir1 = MiniDFSCluster.getRbwDir(storageDir, bpid);
storageDir = cluster.getInstanceStorageDir(sndNode, 1);
File dir2 = MiniDFSCluster.getRbwDir(storageDir, bpid);
while (dir1.listFiles().length != 0 || dir2.listFiles().length != 0) {
Thread.sleep(100);
}
// then increase the file's replication factor
fs.setReplication(fileName, (short) 2);
// replication should succeed
DFSTestUtil.waitReplication(fs, fileName, (short) 1);
// clean up the file
fs.delete(fileName, false);
}