Apache
/
hadoop
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

HDFS-14768. EC : Busy DN replica should be consider in live replica check. Contributed by guojh.

This commit is contained in:
Surendra Singh Lilhore 2019-11-01 20:34:09 +05:30
parent 477505ccfc
commit 02009c3bb7
5 changed files with 113 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

33
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/BlockManager.java
View File

@ -854,7 +854,7 @@ public class BlockManager implements BlockStatsMXBean {
}
// source node returned is not used
chooseSourceDatanodes(blockInfo, containingNodes,
containingLiveReplicasNodes, numReplicas,
containingLiveReplicasNodes, numReplicas, new ArrayList<Byte>(),
new ArrayList<Byte>(), LowRedundancyBlocks.LEVEL);
// containingLiveReplicasNodes can include READ_ONLY_SHARED replicas which are
@ -2024,9 +2024,10 @@ public class BlockManager implements BlockStatsMXBean {
List<DatanodeStorageInfo> liveReplicaNodes = new ArrayList<>();
NumberReplicas numReplicas = new NumberReplicas();
List<Byte> liveBlockIndices = new ArrayList<>();
List<Byte> liveBusyBlockIndices = new ArrayList<>();
final DatanodeDescriptor[] srcNodes = chooseSourceDatanodes(block,
containingNodes, liveReplicaNodes, numReplicas,
liveBlockIndices, priority);
liveBlockIndices, liveBusyBlockIndices, priority);
short requiredRedundancy = getExpectedLiveRedundancyNum(block,
numReplicas);
if(srcNodes == null || srcNodes.length == 0) {
@ -2079,9 +2080,13 @@ public class BlockManager implements BlockStatsMXBean {
for (int i = 0 ; i < liveBlockIndices.size(); i++) {
indices[i] = liveBlockIndices.get(i);
}
byte[] busyIndices = new byte[liveBusyBlockIndices.size()];
for (int i = 0; i < liveBusyBlockIndices.size(); i++) {
busyIndices[i] = liveBusyBlockIndices.get(i);
}
return new ErasureCodingWork(getBlockPoolId(), block, bc, srcNodes,
containingNodes, liveReplicaNodes, additionalReplRequired,
priority, indices);
priority, indices, busyIndices);
} else {
return new ReplicationWork(block, bc, srcNodes,
containingNodes, liveReplicaNodes, additionalReplRequired,
@ -2293,8 +2298,8 @@ public class BlockManager implements BlockStatsMXBean {
DatanodeDescriptor[] chooseSourceDatanodes(BlockInfo block,
List<DatanodeDescriptor> containingNodes,
List<DatanodeStorageInfo> nodesContainingLiveReplicas,
NumberReplicas numReplicas,
List<Byte> liveBlockIndices, int priority) {
NumberReplicas numReplicas, List<Byte> liveBlockIndices,
List<Byte> liveBusyBlockIndices, int priority) {
containingNodes.clear();
nodesContainingLiveReplicas.clear();
List<DatanodeDescriptor> srcNodes = new ArrayList<>();
@ -2347,12 +2352,6 @@ public class BlockManager implements BlockStatsMXBean {
continue;
}
if (priority != LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY
&& (!node.isDecommissionInProgress() && !node.isEnteringMaintenance())
&& node.getNumberOfBlocksToBeReplicated() >= maxReplicationStreams) {
continue; // already reached replication limit
}
// for EC here need to make sure the numReplicas replicates state correct
// because in the scheduleReconstruction it need the numReplicas to check
// whether need to reconstruct the ec internal block
@ -2364,7 +2363,19 @@ public class BlockManager implements BlockStatsMXBean {
liveBitSet, decommissioningBitSet, blockIndex);
}
if (priority != LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY
&& (!node.isDecommissionInProgress() && !node.isEnteringMaintenance())
&& node.getNumberOfBlocksToBeReplicated() >= maxReplicationStreams) {
if (isStriped && state == StoredReplicaState.LIVE) {
liveBusyBlockIndices.add(blockIndex);
}
continue; // already reached replication limit
}
if (node.getNumberOfBlocksToBeReplicated() >= replicationStreamsHardLimit) {
if (isStriped && state == StoredReplicaState.LIVE) {
liveBusyBlockIndices.add(blockIndex);
}
continue;
}

3
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/DatanodeDescriptor.java
View File

@ -634,7 +634,8 @@ public class DatanodeDescriptor extends DatanodeInfo {
return new BlockIterator(startBlock, getStorageInfos());
}
void incrementPendingReplicationWithoutTargets() {
@VisibleForTesting
public void incrementPendingReplicationWithoutTargets() {
pendingReplicationWithoutTargets++;
}

12
hadoop-hdfs-project/hadoop-hdfs/src/main/java/org/apache/hadoop/hdfs/server/blockmanagement/ErasureCodingWork.java
View File

@ -31,6 +31,7 @@ import java.util.Set;
class ErasureCodingWork extends BlockReconstructionWork {
private final byte[] liveBlockIndicies;
private final byte[] liveBusyBlockIndicies;
private final String blockPoolId;
public ErasureCodingWork(String blockPoolId, BlockInfo block,
@ -38,12 +39,13 @@ class ErasureCodingWork extends BlockReconstructionWork {
DatanodeDescriptor[] srcNodes,
List<DatanodeDescriptor> containingNodes,
List<DatanodeStorageInfo> liveReplicaStorages,
int additionalReplRequired,
int priority, byte[] liveBlockIndicies) {
int additionalReplRequired, int priority,
byte[] liveBlockIndicies, byte[] liveBusyBlockIndicies) {
super(block, bc, srcNodes, containingNodes,
liveReplicaStorages, additionalReplRequired, priority);
this.blockPoolId = blockPoolId;
this.liveBlockIndicies = liveBlockIndicies;
this.liveBusyBlockIndicies = liveBusyBlockIndicies;
LOG.debug("Creating an ErasureCodingWork to {} reconstruct ",
block);
}
@ -70,13 +72,17 @@ class ErasureCodingWork extends BlockReconstructionWork {
*/
private boolean hasAllInternalBlocks() {
final BlockInfoStriped block = (BlockInfoStriped) getBlock();
if (getSrcNodes().length < block.getRealTotalBlockNum()) {
if (liveBlockIndicies.length
+ liveBusyBlockIndicies.length < block.getRealTotalBlockNum()) {
return false;
}
BitSet bitSet = new BitSet(block.getTotalBlockNum());
for (byte index : liveBlockIndicies) {
bitSet.set(index);
}
for (byte busyIndex: liveBusyBlockIndicies) {
bitSet.set(busyIndex);
}
for (int i = 0; i < block.getRealDataBlockNum(); i++) {
if (!bitSet.get(i)) {
return false;

68
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/TestDecommissionWithStriped.java
View File

@ -40,14 +40,16 @@ import org.apache.hadoop.fs.FileSystem;
import org.apache.hadoop.fs.Path;
import org.apache.hadoop.hdfs.MiniDFSCluster.DataNodeProperties;
import org.apache.hadoop.hdfs.client.HdfsDataInputStream;
import org.apache.hadoop.hdfs.protocol.Block;
import org.apache.hadoop.hdfs.protocol.DatanodeInfo;
import org.apache.hadoop.hdfs.protocol.DatanodeInfo.AdminStates;
import org.apache.hadoop.hdfs.protocol.ErasureCodingPolicy;
import org.apache.hadoop.hdfs.protocol.HdfsConstants.DatanodeReportType;
import org.apache.hadoop.hdfs.protocol.LocatedBlocks;
import org.apache.hadoop.hdfs.security.token.block.BlockTokenIdentifier;
import org.apache.hadoop.hdfs.protocol.Block;
import org.apache.hadoop.hdfs.protocol.LocatedBlock;
import org.apache.hadoop.hdfs.protocol.LocatedStripedBlock;
import org.apache.hadoop.hdfs.server.blockmanagement.BlockInfo;
import org.apache.hadoop.hdfs.server.blockmanagement.BlockInfoStriped;
import org.apache.hadoop.hdfs.server.blockmanagement.BlockManager;
import org.apache.hadoop.hdfs.server.blockmanagement.BlockManagerTestUtil;
@ -55,6 +57,7 @@ import org.apache.hadoop.hdfs.server.blockmanagement.DatanodeDescriptor;
import org.apache.hadoop.hdfs.server.blockmanagement.DatanodeStorageInfo;
import org.apache.hadoop.hdfs.server.datanode.DataNode;
import org.apache.hadoop.hdfs.server.namenode.FSNamesystem;
import org.apache.hadoop.hdfs.server.namenode.INodeFile;
import org.apache.hadoop.hdfs.server.namenode.NameNode;
import org.apache.hadoop.hdfs.server.namenode.NameNodeAdapter;
import org.apache.hadoop.security.token.Token;
@ -81,6 +84,9 @@ public class TestDecommissionWithStriped {
// replication interval
private static final int NAMENODE_REPLICATION_INTERVAL = 1;
private int replicationStreamsHardLimit =
DFSConfigKeys.DFS_NAMENODE_REPLICATION_STREAMS_HARD_LIMIT_DEFAULT;
private Path decommissionDir;
private Path hostsFile;
private Path excludeFile;
@ -273,7 +279,6 @@ public class TestDecommissionWithStriped {
fsn.getNumDecomLiveDataNodes());
// Ensure decommissioned datanode is not automatically shutdown
DFSClient client = getDfsClient(cluster.getNameNode(0), conf);
assertEquals("All datanodes must be alive", numDNs,
client.datanodeReport(DatanodeReportType.LIVE).length);
@ -283,6 +288,65 @@ public class TestDecommissionWithStriped {
cleanupFile(dfs, ecFile);
}
/**
* DN decommission shouldn't reconstruction busy DN block.
* @throws Exception
*/
@Test(timeout = 120000)
public void testDecommissionWithBusyNode() throws Exception {
byte busyDNIndex = 1;
byte decommisionDNIndex = 0;
//1. create EC file
final Path ecFile = new Path(ecDir, "testDecommissionWithBusyNode");
int writeBytes = cellSize * dataBlocks;
writeStripedFile(dfs, ecFile, writeBytes);
Assert.assertEquals(0, bm.numOfUnderReplicatedBlocks());
FileChecksum fileChecksum1 = dfs.getFileChecksum(ecFile, writeBytes);
//2. make once DN busy
final INodeFile fileNode = cluster.getNamesystem().getFSDirectory()
.getINode4Write(ecFile.toString()).asFile();
BlockInfo firstBlock = fileNode.getBlocks()[0];
DatanodeStorageInfo[] dnStorageInfos = bm.getStorages(firstBlock);
DatanodeDescriptor busyNode =
dnStorageInfos[busyDNIndex].getDatanodeDescriptor();
for (int j = 0; j < replicationStreamsHardLimit; j++) {
busyNode.incrementPendingReplicationWithoutTargets();
}
//3. decomission one node
List<DatanodeInfo> decommisionNodes = new ArrayList<>();
decommisionNodes.add(
dnStorageInfos[decommisionDNIndex].getDatanodeDescriptor());
decommissionNode(0, decommisionNodes, AdminStates.DECOMMISSIONED);
assertEquals(decommisionNodes.size(), fsn.getNumDecomLiveDataNodes());
//4. wait for decommission block to replicate
Thread.sleep(3000);
DatanodeStorageInfo[] newDnStorageInfos = bm.getStorages(firstBlock);
Assert.assertEquals("Busy DN shouldn't be reconstructed",
dnStorageInfos[busyDNIndex].getStorageID(),
newDnStorageInfos[busyDNIndex].getStorageID());
//5. check decommission DN block index, it should be reconstructed again
LocatedBlocks lbs = cluster.getNameNodeRpc().getBlockLocations(
ecFile.toString(), 0, writeBytes);
LocatedStripedBlock bg = (LocatedStripedBlock) (lbs.get(0));
int decommissionBlockIndexCount = 0;
for (byte index : bg.getBlockIndices()) {
if (index == decommisionDNIndex) {
decommissionBlockIndexCount++;
}
}
Assert.assertEquals("Decommission DN block should be reconstructed", 2,
decommissionBlockIndexCount);
FileChecksum fileChecksum2 = dfs.getFileChecksum(ecFile, writeBytes);
Assert.assertTrue("Checksum mismatches!",
fileChecksum1.equals(fileChecksum2));
}
/**
* Tests to verify that the file checksum should be able to compute after the
* decommission operation.

17
hadoop-hdfs-project/hadoop-hdfs/src/test/java/org/apache/hadoop/hdfs/server/blockmanagement/TestBlockManager.java
View File

@ -660,6 +660,7 @@ public class TestBlockManager {
liveNodes,
new NumberReplicas(),
new ArrayList<Byte>(),
new ArrayList<Byte>(),
LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY)[0]);
assertEquals("Does not choose a source node for a less-than-highest-priority"
@ -671,6 +672,7 @@ public class TestBlockManager {
liveNodes,
new NumberReplicas(),
new ArrayList<Byte>(),
new ArrayList<Byte>(),
LowRedundancyBlocks.QUEUE_VERY_LOW_REDUNDANCY).length);
// Increase the replication count to test replication count > hard limit
@ -685,6 +687,7 @@ public class TestBlockManager {
liveNodes,
new NumberReplicas(),
new ArrayList<Byte>(),
new ArrayList<Byte>(),
LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY).length);
}
@ -730,13 +733,15 @@ public class TestBlockManager {
List<DatanodeStorageInfo> liveNodes = new LinkedList<DatanodeStorageInfo>();
NumberReplicas numReplicas = new NumberReplicas();
List<Byte> liveBlockIndices = new ArrayList<>();
List<Byte> liveBusyBlockIndices = new ArrayList<>();
bm.chooseSourceDatanodes(
aBlockInfoStriped,
cntNodes,
liveNodes,
numReplicas, liveBlockIndices,
LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY);
liveBusyBlockIndices,
LowRedundancyBlocks.QUEUE_VERY_LOW_REDUNDANCY);
assertEquals("Choose the source node for reconstruction with one node reach"
+ " the MAX maxReplicationStreams, the numReplicas still return the"
@ -791,12 +796,14 @@ public class TestBlockManager {
new LinkedList<DatanodeStorageInfo>();
NumberReplicas numReplicas = new NumberReplicas();
List<Byte> liveBlockIndices = new ArrayList<>();
List<Byte> liveBusyBlockIndices = new ArrayList<>();
bm.chooseSourceDatanodes(
aBlockInfoStriped,
containingNodes,
nodesContainingLiveReplicas,
numReplicas, liveBlockIndices,
liveBusyBlockIndices,
LowRedundancyBlocks.QUEUE_HIGHEST_PRIORITY);
assertEquals("There are 5 live replicas in " +
"[ds2, ds3, ds4, ds5, ds6] datanodes ",
@ -828,7 +835,9 @@ public class TestBlockManager {
bm.getStoredBlock(aBlock),
cntNodes,
liveNodes,
new NumberReplicas(), new LinkedList<Byte>(),
new NumberReplicas(),
new LinkedList<Byte>(),
new ArrayList<Byte>(),
LowRedundancyBlocks.QUEUE_LOW_REDUNDANCY)[0]);
@ -842,7 +851,9 @@ public class TestBlockManager {
bm.getStoredBlock(aBlock),
cntNodes,
liveNodes,
new NumberReplicas(), new LinkedList<Byte>(),
new NumberReplicas(),
new LinkedList<Byte>(),
new ArrayList<Byte>(),
LowRedundancyBlocks.QUEUE_LOW_REDUNDANCY).length);
}