HDFS-9866. BlockManager#chooseExcessReplicasStriped may weaken rack fault tolerance. Contributed by Jing Zhao.

This commit is contained in:
Jing Zhao 2016-02-28 14:54:49 -08:00
parent 7634d404b7
commit 408f2c807b
9 changed files with 133 additions and 112 deletions

View File

@ -446,6 +446,9 @@ Trunk (Unreleased)
HDFS-9734. Refactoring of checksum failure report related codes.
(Kai Zheng via zhz)
HDFS-9866. BlockManager#chooseExcessReplicasStriped may weaken rack fault
tolerance. (jing9)
BREAKDOWN OF HDFS-7285 SUBTASKS AND RELATED JIRAS
HDFS-7347. Configurable erasure coding policy for individual files and

View File

@ -3245,7 +3245,7 @@ private void chooseExcessReplicasContiguous(
DatanodeDescriptor delNodeHint, List<StorageType> excessTypes) {
BlockPlacementPolicy replicator = placementPolicies.getPolicy(false);
List<DatanodeStorageInfo> replicasToDelete = replicator
.chooseReplicasToDelete(nonExcess, replication, excessTypes,
.chooseReplicasToDelete(nonExcess, nonExcess, replication, excessTypes,
addedNode, delNodeHint);
for (DatanodeStorageInfo choosenReplica : replicasToDelete) {
processChosenExcessReplica(nonExcess, choosenReplica, storedBlock);
@ -3316,8 +3316,8 @@ private void chooseExcessReplicasStriped(BlockCollection bc,
internalBlock.setBlockId(storedBlock.getBlockId() + targetIndex);
while (candidates.size() > 1) {
List<DatanodeStorageInfo> replicasToDelete = placementPolicy
.chooseReplicasToDelete(candidates, (short) 1, excessTypes, null,
null);
.chooseReplicasToDelete(nonExcess, candidates, (short) 1,
excessTypes, null, null);
for (DatanodeStorageInfo chosen : replicasToDelete) {
processChosenExcessReplica(nonExcess, chosen, storedBlock);
candidates.remove(chosen);

View File

@ -77,8 +77,6 @@ public abstract DatanodeStorageInfo[] chooseTarget(String srcPath,
BlockStoragePolicy storagePolicy);
/**
* Same as {@link #chooseTarget(String, int, Node, Set, long, List, StorageType)}
* with added parameter {@code favoredDatanodes}
* @param favoredNodes datanodes that should be favored as targets. This
* is only a hint and due to cluster state, namenode may not be
* able to place the blocks on these datanodes.
@ -106,17 +104,21 @@ DatanodeStorageInfo[] chooseTarget(String src,
* @param numOfReplicas replica number of file to be verified
* @return the result of verification
*/
abstract public BlockPlacementStatus verifyBlockPlacement(
public abstract BlockPlacementStatus verifyBlockPlacement(
DatanodeInfo[] locs, int numOfReplicas);
/**
* Select the excess replica storages for deletion based on either
* delNodehint/Excess storage types.
*
* @param candidates
* @param availableReplicas
* available replicas
* @param delCandidates
* Candidates for deletion. For normal replication, this set is the
* same with availableReplicas. For striped blocks, this set is a
* subset of availableReplicas.
* @param expectedNumOfReplicas
* The required number of replicas for this block
* The expected number of replicas remaining in the delCandidates
* @param excessTypes
* type of the storagepolicy
* @param addedNode
@ -125,10 +127,12 @@ abstract public BlockPlacementStatus verifyBlockPlacement(
* Hint for excess storage selection
* @return Returns the list of excess replicas chosen for deletion
*/
abstract public List<DatanodeStorageInfo> chooseReplicasToDelete(
Collection<DatanodeStorageInfo> candidates, int expectedNumOfReplicas,
public abstract List<DatanodeStorageInfo> chooseReplicasToDelete(
Collection<DatanodeStorageInfo> availableReplicas,
Collection<DatanodeStorageInfo> delCandidates, int expectedNumOfReplicas,
List<StorageType> excessTypes, DatanodeDescriptor addedNode,
DatanodeDescriptor delNodeHint);
/**
* Used to setup a BlockPlacementPolicy object. This should be defined by
* all implementations of a BlockPlacementPolicy.
@ -137,7 +141,7 @@ abstract public List<DatanodeStorageInfo> chooseReplicasToDelete(
* @param stats retrieve cluster status from here
* @param clusterMap cluster topology
*/
abstract protected void initialize(Configuration conf, FSClusterStats stats,
protected abstract void initialize(Configuration conf, FSClusterStats stats,
NetworkTopology clusterMap,
Host2NodesMap host2datanodeMap);
@ -149,7 +153,7 @@ abstract protected void initialize(Configuration conf, FSClusterStats stats,
* @param source source replica of the move
* @param target target replica of the move
*/
abstract public boolean isMovable(Collection<DatanodeInfo> candidates,
public abstract boolean isMovable(Collection<DatanodeInfo> candidates,
DatanodeInfo source, DatanodeInfo target);
/**
@ -191,10 +195,8 @@ protected <T> DatanodeInfo getDatanodeInfo(T datanode) {
"class " + datanode.getClass().getName() + " not allowed");
if (datanode instanceof DatanodeInfo) {
return ((DatanodeInfo)datanode);
} else if (datanode instanceof DatanodeStorageInfo) {
return ((DatanodeStorageInfo)datanode).getDatanodeDescriptor();
} else {
return null;
return ((DatanodeStorageInfo)datanode).getDatanodeDescriptor();
}
}
@ -209,35 +211,37 @@ protected String getRack(final DatanodeInfo datanode) {
/**
* Split data nodes into two sets, one set includes nodes on rack with
* more than one replica, the other set contains the remaining nodes.
*
* @param storagesOrDataNodes DatanodeStorageInfo/DatanodeInfo to be split
*
* @param availableSet all the available DataNodes/storages of the block
* @param candidates DatanodeStorageInfo/DatanodeInfo to be split
* into two sets
* @param rackMap a map from rack to datanodes
* @param moreThanOne contains nodes on rack with more than one replica
* @param exactlyOne remains contains the remaining nodes
*/
public <T> void splitNodesWithRack(
final Iterable<T> storagesOrDataNodes,
final Iterable<T> availableSet,
final Collection<T> candidates,
final Map<String, List<T>> rackMap,
final List<T> moreThanOne,
final List<T> exactlyOne) {
for(T s: storagesOrDataNodes) {
for(T s: availableSet) {
final String rackName = getRack(getDatanodeInfo(s));
List<T> storageList = rackMap.get(rackName);
if (storageList == null) {
storageList = new ArrayList<T>();
storageList = new ArrayList<>();
rackMap.put(rackName, storageList);
}
storageList.add(s);
}
// split nodes into two sets
for(List<T> storageList : rackMap.values()) {
if (storageList.size() == 1) {
for (T candidate : candidates) {
final String rackName = getRack(getDatanodeInfo(candidate));
if (rackMap.get(rackName).size() == 1) {
// exactlyOne contains nodes on rack with only one replica
exactlyOne.add(storageList.get(0));
exactlyOne.add(candidate);
} else {
// moreThanOne contains nodes on rack with more than one replica
moreThanOne.addAll(storageList);
moreThanOne.add(candidate);
}
}
}

View File

@ -972,7 +972,8 @@ public DatanodeStorageInfo chooseReplicaToDelete(
@Override
public List<DatanodeStorageInfo> chooseReplicasToDelete(
Collection<DatanodeStorageInfo> candidates,
Collection<DatanodeStorageInfo> availableReplicas,
Collection<DatanodeStorageInfo> delCandidates,
int expectedNumOfReplicas,
List<StorageType> excessTypes,
DatanodeDescriptor addedNode,
@ -985,28 +986,29 @@ public List<DatanodeStorageInfo> chooseReplicasToDelete(
final List<DatanodeStorageInfo> moreThanOne = new ArrayList<>();
final List<DatanodeStorageInfo> exactlyOne = new ArrayList<>();
// split nodes into two sets
// split candidate nodes for deletion into two sets
// moreThanOne contains nodes on rack with more than one replica
// exactlyOne contains the remaining nodes
splitNodesWithRack(candidates, rackMap, moreThanOne, exactlyOne);
splitNodesWithRack(availableReplicas, delCandidates, rackMap, moreThanOne,
exactlyOne);
// pick one node to delete that favors the delete hint
// otherwise pick one with least space from priSet if it is not empty
// otherwise one node with least space from remains
boolean firstOne = true;
final DatanodeStorageInfo delNodeHintStorage =
DatanodeStorageInfo.getDatanodeStorageInfo(candidates, delNodeHint);
DatanodeStorageInfo.getDatanodeStorageInfo(delCandidates, delNodeHint);
final DatanodeStorageInfo addedNodeStorage =
DatanodeStorageInfo.getDatanodeStorageInfo(candidates, addedNode);
DatanodeStorageInfo.getDatanodeStorageInfo(delCandidates, addedNode);
while (candidates.size() - expectedNumOfReplicas > excessReplicas.size()) {
while (delCandidates.size() - expectedNumOfReplicas > excessReplicas.size()) {
final DatanodeStorageInfo cur;
if (firstOne && useDelHint(delNodeHintStorage, addedNodeStorage,
moreThanOne, exactlyOne, excessTypes)) {
cur = delNodeHintStorage;
} else { // regular excessive replica removal
cur = chooseReplicaToDelete(moreThanOne, exactlyOne, excessTypes,
rackMap);
cur = chooseReplicaToDelete(moreThanOne, exactlyOne,
excessTypes, rackMap);
}
firstOne = false;
if (cur == null) {
@ -1056,7 +1058,7 @@ public boolean isMovable(Collection<DatanodeInfo> locs,
final Map<String, List<DatanodeInfo>> rackMap = new HashMap<>();
final List<DatanodeInfo> moreThanOne = new ArrayList<>();
final List<DatanodeInfo> exactlyOne = new ArrayList<>();
splitNodesWithRack(locs, rackMap, moreThanOne, exactlyOne);
splitNodesWithRack(locs, locs, rackMap, moreThanOne, exactlyOne);
return notReduceNumOfGroups(moreThanOne, source, target);
}

View File

@ -23,6 +23,7 @@
import org.apache.hadoop.net.Node;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.BitSet;
import java.util.HashMap;
import java.util.List;
@ -137,6 +138,11 @@ void addTaskToDatanode() {
stripedBlk.getBlockId() + blockIndex, internBlkLen,
stripedBlk.getGenerationStamp());
source.addBlockToBeReplicated(targetBlk, getTargets());
if (BlockManager.LOG.isDebugEnabled()) {
BlockManager.LOG.debug("Add replication task from source {} to " +
"targets {} for EC block {}", source, Arrays.toString(getTargets()),
targetBlk);
}
} else {
getTargets()[0].getDatanodeDescriptor().addBlockToBeErasureCoded(
new ExtendedBlock(blockPoolId, stripedBlk),

View File

@ -23,8 +23,9 @@
import org.apache.hadoop.hdfs.DistributedFileSystem;
import org.apache.hadoop.hdfs.HdfsConfiguration;
import org.apache.hadoop.hdfs.MiniDFSCluster;
import org.apache.hadoop.hdfs.protocol.Block;
import org.apache.hadoop.hdfs.protocol.ExtendedBlock;
import org.apache.hadoop.hdfs.protocol.DatanodeInfo;
import org.apache.hadoop.hdfs.protocol.LocatedBlocks;
import org.apache.hadoop.hdfs.protocol.LocatedStripedBlock;
import org.apache.hadoop.hdfs.server.datanode.DataNode;
import org.apache.hadoop.hdfs.server.datanode.DataNodeTestUtils;
import org.apache.hadoop.hdfs.server.namenode.FSNamesystem;
@ -39,9 +40,8 @@
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import java.util.Arrays;
import java.io.IOException;
import java.util.HashSet;
import java.util.List;
import java.util.Set;
import static org.apache.hadoop.hdfs.StripedFileTestUtil.BLOCK_STRIPED_CELL_SIZE;
@ -55,14 +55,13 @@ public class TestReconstructStripedBlocksWithRackAwareness {
static {
GenericTestUtils.setLogLevel(BlockPlacementPolicy.LOG, Level.ALL);
GenericTestUtils.setLogLevel(BlockManager.blockLog, Level.ALL);
GenericTestUtils.setLogLevel(BlockManager.LOG, Level.ALL);
}
private static final String[] hosts = new String[]{"host1", "host2", "host3",
"host4", "host5", "host6", "host7", "host8", "host9", "host10"};
private static final String[] racks = new String[]{"/r1", "/r1", "/r2", "/r2",
"/r3", "/r3", "/r4", "/r4", "/r5", "/r6"};
private static final List<String> singleNodeRacks = Arrays.asList("host9", "host10");
private static final short blockNum = (short) (NUM_DATA_BLOCKS + NUM_PARITY_BLOCKS);
private MiniDFSCluster cluster;
private DistributedFileSystem fs;
@ -94,6 +93,20 @@ public void tearDown() {
}
}
private MiniDFSCluster.DataNodeProperties stopDataNode(String hostname)
throws IOException {
MiniDFSCluster.DataNodeProperties dnProp = null;
for (int i = 0; i < cluster.getDataNodes().size(); i++) {
DataNode dn = cluster.getDataNodes().get(i);
if (dn.getDatanodeId().getHostName().equals(hostname)) {
dnProp = cluster.stopDataNode(i);
cluster.setDataNodeDead(dn.getDatanodeId());
LOG.info("stop datanode " + dn.getDatanodeId().getHostName());
}
}
return dnProp;
}
/**
* When there are all the internal blocks available but they are not placed on
* enough racks, NameNode should avoid normal decoding reconstruction but copy
@ -102,24 +115,13 @@ public void tearDown() {
* In this test, we first need to create a scenario that a striped block has
* all the internal blocks but distributed in <6 racks. Then we check if the
* replication monitor can correctly schedule the reconstruction work for it.
*
* For the 9 internal blocks + 5 racks setup, the test does the following:
* 1. create a 6 rack cluster with 10 datanodes, where there are 2 racks only
* containing 1 datanodes each
* 2. for a striped block with 9 internal blocks, there must be one internal
* block locating in a single-node rack. find this node and stop it
* 3. namenode will trigger reconstruction for the block and since the cluster
* has only 5 racks remaining, after the reconstruction we have 9 internal
* blocks distributed in 5 racks.
* 4. we bring the datanode back, now the cluster has 6 racks again
* 5. let the datanode call reportBadBlock, this will make the namenode to
* check if the striped block is placed in >= 6 racks, and the namenode will
* put the block into the under-replicated queue
* 6. now we can check if the replication monitor works as expected
*/
@Test
public void testReconstructForNotEnoughRacks() throws Exception {
MiniDFSCluster.DataNodeProperties host10 = stopDataNode("host10");
final Path file = new Path("/foo");
// the file's block is in 9 dn but 5 racks
DFSTestUtil.createFile(fs, file,
BLOCK_STRIPED_CELL_SIZE * NUM_DATA_BLOCKS * 2, (short) 1, 0L);
Assert.assertEquals(0, bm.numOfUnderReplicatedBlocks());
@ -128,39 +130,6 @@ public void testReconstructForNotEnoughRacks() throws Exception {
.getINode4Write(file.toString()).asFile();
BlockInfoStriped blockInfo = (BlockInfoStriped) fileNode.getLastBlock();
// find the internal block located in the single node rack
Block internalBlock = null;
String hostToStop = null;
for (DatanodeStorageInfo storage : blockInfo.storages) {
if (singleNodeRacks.contains(storage.getDatanodeDescriptor().getHostName())) {
hostToStop = storage.getDatanodeDescriptor().getHostName();
internalBlock = blockInfo.getBlockOnStorage(storage);
}
}
Assert.assertNotNull(internalBlock);
Assert.assertNotNull(hostToStop);
// delete the block on the chosen datanode
cluster.corruptBlockOnDataNodesByDeletingBlockFile(
new ExtendedBlock(bm.getBlockPoolId(), internalBlock));
// stop the chosen datanode
MiniDFSCluster.DataNodeProperties dnProp = null;
for (int i = 0; i < cluster.getDataNodes().size(); i++) {
DataNode dn = cluster.getDataNodes().get(i);
if (dn.getDatanodeId().getHostName().equals(hostToStop)) {
dnProp = cluster.stopDataNode(i);
cluster.setDataNodeDead(dn.getDatanodeId());
LOG.info("stop datanode " + dn.getDatanodeId().getHostName());
}
}
NetworkTopology topology = bm.getDatanodeManager().getNetworkTopology();
Assert.assertEquals(5, topology.getNumOfRacks());
// make sure the reconstruction work can finish
// now we have 9 internal blocks in 5 racks
DFSTestUtil.waitForReplication(fs, file, blockNum, 15 * 1000);
// we now should have 9 internal blocks distributed in 5 racks
Set<String> rackSet = new HashSet<>();
for (DatanodeStorageInfo storage : blockInfo.storages) {
@ -169,27 +138,25 @@ public void testReconstructForNotEnoughRacks() throws Exception {
Assert.assertEquals(5, rackSet.size());
// restart the stopped datanode
cluster.restartDataNode(dnProp);
cluster.restartDataNode(host10);
cluster.waitActive();
// make sure we have 6 racks again
topology = bm.getDatanodeManager().getNetworkTopology();
NetworkTopology topology = bm.getDatanodeManager().getNetworkTopology();
Assert.assertEquals(hosts.length, topology.getNumOfLeaves());
Assert.assertEquals(6, topology.getNumOfRacks());
// pause all the heartbeats
DataNode badDn = null;
for (DataNode dn : cluster.getDataNodes()) {
DataNodeTestUtils.setHeartbeatsDisabledForTests(dn, true);
if (dn.getDatanodeId().getHostName().equals(hostToStop)) {
badDn = dn;
}
}
assert badDn != null;
// let the DN report the bad block, so that the namenode will put the block
// into under-replicated queue. note that the block still has 9 internal
// blocks but in 5 racks
badDn.reportBadBlocks(new ExtendedBlock(bm.getBlockPoolId(), internalBlock));
fsn.writeLock();
try {
bm.processMisReplicatedBlocks();
} finally {
fsn.writeUnlock();
}
// check if replication monitor correctly schedule the replication work
boolean scheduled = false;
@ -210,4 +177,42 @@ public void testReconstructForNotEnoughRacks() throws Exception {
}
Assert.assertTrue(scheduled);
}
@Test
public void testChooseExcessReplicasToDelete() throws Exception {
MiniDFSCluster.DataNodeProperties host10 = stopDataNode("host10");
final Path file = new Path("/foo");
DFSTestUtil.createFile(fs, file,
BLOCK_STRIPED_CELL_SIZE * NUM_DATA_BLOCKS * 2, (short) 1, 0L);
// stop host1
MiniDFSCluster.DataNodeProperties host1 = stopDataNode("host1");
// bring host10 back
cluster.restartDataNode(host10);
cluster.waitActive();
// wait for reconstruction to finish
final short blockNum = (short) (NUM_DATA_BLOCKS + NUM_PARITY_BLOCKS);
DFSTestUtil.waitForReplication(fs, file, blockNum, 15 * 1000);
// restart host1
cluster.restartDataNode(host1);
cluster.waitActive();
for (DataNode dn : cluster.getDataNodes()) {
if (dn.getDatanodeId().getHostName().equals("host1")) {
DataNodeTestUtils.triggerBlockReport(dn);
break;
}
}
// make sure the excess replica is detected, and we delete host1's replica
// so that we have 6 racks
DFSTestUtil.waitForReplication(fs, file, blockNum, 15 * 1000);
LocatedBlocks blks = fs.getClient().getLocatedBlocks(file.toString(), 0);
LocatedStripedBlock block = (LocatedStripedBlock) blks.getLastLocatedBlock();
for (DatanodeInfo dn : block.getLocations()) {
Assert.assertFalse(dn.getHostName().equals("host1"));
}
}
}

View File

@ -965,7 +965,8 @@ public void testChooseReplicaToDelete() throws Exception {
List<DatanodeStorageInfo> first = new ArrayList<>();
List<DatanodeStorageInfo> second = new ArrayList<>();
replicator.splitNodesWithRack(replicaList, rackMap, first, second);
replicator.splitNodesWithRack(replicaList, replicaList, rackMap, first,
second);
// storages[0] and storages[1] are in first set as their rack has two
// replica nodes, while storages[2] and dataNodes[5] are in second set.
assertEquals(2, first.size());
@ -1018,7 +1019,7 @@ public void testChooseReplicasToDelete() throws Exception {
DatanodeDescriptor delHintNode = storages[0].getDatanodeDescriptor();
List<StorageType> excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(), delHintNode);
assertTrue(excessReplicas.size() == 1);
assertTrue(excessReplicas.contains(storages[0]));
@ -1031,7 +1032,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(excessStorage);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(), null);
assertTrue(excessReplicas.contains(excessStorage));
@ -1051,7 +1052,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[5]);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(),
storages[5].getDatanodeDescriptor());
assertEquals(1, excessReplicas.size());
@ -1070,7 +1071,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[3]);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[1].getDatanodeDescriptor(),
storages[3].getDatanodeDescriptor());
assertEquals(1, excessReplicas.size());
@ -1084,7 +1085,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[2]);
excessTypes = storagePolicy.chooseExcess((short) 1,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 1,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 1,
excessTypes, storages[2].getDatanodeDescriptor(), null);
assertEquals(1, excessReplicas.size());
assertTrue(excessReplicas.contains(excessSSD));
@ -1104,7 +1105,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[5]);
excessTypes = storagePolicy.chooseExcess((short) 2,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 2,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 2,
excessTypes, null, null);
assertEquals(0, excessReplicas.size());
}

View File

@ -637,7 +637,7 @@ public void testChooseReplicaToDelete() throws Exception {
List<DatanodeStorageInfo> first = new ArrayList<>();
List<DatanodeStorageInfo> second = new ArrayList<>();
replicator.splitNodesWithRack(
replicator.splitNodesWithRack(replicaList,
replicaList, rackMap, first, second);
assertEquals(3, first.size());
assertEquals(1, second.size());

View File

@ -330,7 +330,7 @@ public void testChooseReplicasToDelete() throws Exception {
DatanodeDescriptor delHintNode = storages[0].getDatanodeDescriptor();
List<StorageType> excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(), delHintNode);
assertTrue(excessReplicas.size() == 1);
assertTrue(excessReplicas.contains(storages[0]));
@ -340,7 +340,7 @@ public void testChooseReplicasToDelete() throws Exception {
delHintNode = storages[1].getDatanodeDescriptor();
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(), delHintNode);
assertTrue(excessReplicas.size() == 1);
assertTrue(excessReplicas.contains(storages[0]));
@ -353,7 +353,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[8]);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[8].getDatanodeDescriptor(), null);
assertTrue(excessReplicas.size() == 1);
assertTrue(excessReplicas.contains(storages[1]));
@ -366,7 +366,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[5]);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[8].getDatanodeDescriptor(), null);
assertTrue(excessReplicas.size() == 1);
assertTrue(excessReplicas.contains(storages[1]) ||
@ -384,7 +384,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(excessStorage);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(), null);
assertTrue(excessReplicas.size() == 2);
assertTrue(excessReplicas.contains(storages[0]));
@ -416,7 +416,7 @@ public void testChooseReplicasToDelete() throws Exception {
nonExcess.add(storages[8]);
excessTypes = storagePolicy.chooseExcess((short) 3,
DatanodeStorageInfo.toStorageTypes(nonExcess));
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, 3,
excessReplicas = replicator.chooseReplicasToDelete(nonExcess, nonExcess, 3,
excessTypes, storages[3].getDatanodeDescriptor(),
storages[7].getDatanodeDescriptor());
assertEquals(1, excessReplicas.size());