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

HBASE-15065 SimpleRegionNormalizer should return multiple normalization plans in one run

This commit is contained in:
tedyu 2016-01-08 11:47:34 -08:00
parent 164aeb5399
commit 6e781a1f34
6 changed files with 82 additions and 78 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
hbase-server/src/main/java/org/apache/hadoop/hbase/master/HMaster.java
View File

@ -1341,12 +1341,16 @@ public class HMaster extends HRegionServer implements MasterServices {
continue;
}
}
NormalizationPlan plan = this.normalizer.computePlanForTable(table, types);
plan.execute(clusterConnection.getAdmin());
if (plan.getType() == PlanType.SPLIT) {
splitPlanCount++;
} else if (plan.getType() == PlanType.MERGE) {
mergePlanCount++;
List<NormalizationPlan> plans = this.normalizer.computePlanForTable(table, types);
if (plans != null) {
for (NormalizationPlan plan : plans) {
plan.execute(clusterConnection.getAdmin());
if (plan.getType() == PlanType.SPLIT) {
splitPlanCount++;
} else if (plan.getType() == PlanType.MERGE) {
mergePlanCount++;
}
}
}
}
}

4
hbase-server/src/main/java/org/apache/hadoop/hbase/master/normalizer/RegionNormalizer.java
View File

@ -51,9 +51,9 @@ public interface RegionNormalizer {
* Computes next optimal normalization plan.
* @param table table to normalize
* @param types desired types of NormalizationPlan
* @return Next (perhaps most urgent) normalization action to perform
* @return normalization actions to perform. Null if no action to take
*/
NormalizationPlan computePlanForTable(TableName table, List<PlanType> types)
List<NormalizationPlan> computePlanForTable(TableName table, List<PlanType> types)
throws HBaseIOException;
/**

2
hbase-server/src/main/java/org/apache/hadoop/hbase/master/normalizer/RegionNormalizerChore.java
View File

@ -38,7 +38,7 @@ public class RegionNormalizerChore extends ScheduledChore {
public RegionNormalizerChore(HMaster master) {
super(master.getServerName() + "-RegionNormalizerChore", master,
master.getConfiguration().getInt("hbase.normalizer.period", 1800000));
master.getConfiguration().getInt("hbase.normalizer.period", 300000));
this.master = master;
}

95
hbase-server/src/main/java/org/apache/hadoop/hbase/master/normalizer/SimpleRegionNormalizer.java
View File

@ -72,22 +72,6 @@ public class SimpleRegionNormalizer implements RegionNormalizer {
this.masterServices = masterServices;
}
/*
* This comparator compares the region size.
* The second element in the triple is region size while the 3rd element
* is the index of the region in the underlying List
*/
private Comparator<Triple<HRegionInfo, Long, Integer>> regionSizeComparator =
new Comparator<Triple<HRegionInfo, Long, Integer>>() {
@Override
public int compare(Triple<HRegionInfo, Long, Integer> pair,
Triple<HRegionInfo, Long, Integer> pair2) {
long sz = pair.getSecond();
long sz2 = pair2.getSecond();
return (sz < sz2) ? -1 : ((sz == sz2) ? 0 : 1);
}
};
@Override
public void planSkipped(HRegionInfo hri, PlanType type) {
skippedCount[type.ordinal()]++;
@ -98,6 +82,17 @@ public class SimpleRegionNormalizer implements RegionNormalizer {
return skippedCount[type.ordinal()];
}
// Comparator that gives higher priority to region Split plan
private Comparator<NormalizationPlan> planComparator =
new Comparator<NormalizationPlan>() {
@Override
public int compare(NormalizationPlan plan, NormalizationPlan plan2) {
if (plan instanceof SplitNormalizationPlan) return -1;
if (plan2 instanceof SplitNormalizationPlan) return 1;
return 0;
}
};
/**
* Computes next most "urgent" normalization action on the table.
* Action may be either a split, or a merge, or no action.
@ -107,13 +102,14 @@ public class SimpleRegionNormalizer implements RegionNormalizer {
* @return normalization plan to execute
*/
@Override
public NormalizationPlan computePlanForTable(TableName table, List<PlanType> types)
public List<NormalizationPlan> computePlanForTable(TableName table, List<PlanType> types)
throws HBaseIOException {
if (table == null || table.isSystemTable()) {
LOG.debug("Normalization of system table " + table + " isn't allowed");
return EmptyNormalizationPlan.getInstance();
return null;
}
List<NormalizationPlan> plans = new ArrayList<NormalizationPlan>();
List<HRegionInfo> tableRegions = masterServices.getAssignmentManager().getRegionStates().
getRegionsOfTable(table);
@ -122,7 +118,7 @@ public class SimpleRegionNormalizer implements RegionNormalizer {
int nrRegions = tableRegions == null ? 0 : tableRegions.size();
LOG.debug("Table " + table + " has " + nrRegions + " regions, required min number"
+ " of regions for normalizer to run is " + MIN_REGION_COUNT + ", not running normalizer");
return EmptyNormalizationPlan.getInstance();
return null;
}
LOG.debug("Computing normalization plan for table: " + table +
@ -130,56 +126,49 @@ public class SimpleRegionNormalizer implements RegionNormalizer {
long totalSizeMb = 0;
ArrayList<Triple<HRegionInfo, Long, Integer>> regionsWithSize =
new ArrayList<Triple<HRegionInfo, Long, Integer>>(tableRegions.size());
for (int i = 0; i < tableRegions.size(); i++) {
HRegionInfo hri = tableRegions.get(i);
long regionSize = getRegionSize(hri);
regionsWithSize.add(new Triple<HRegionInfo, Long, Integer>(hri, regionSize, i));
totalSizeMb += regionSize;
}
Collections.sort(regionsWithSize, regionSizeComparator);
Triple<HRegionInfo, Long, Integer> largestRegion = regionsWithSize.get(tableRegions.size()-1);
double avgRegionSize = totalSizeMb / (double) tableRegions.size();
LOG.debug("Table " + table + ", total aggregated regions size: " + totalSizeMb);
LOG.debug("Table " + table + ", average region size: " + avgRegionSize);
// now; if the largest region is >2 times large than average, we split it, split
// is more high priority normalization action than merge.
if (types.contains(PlanType.SPLIT) && largestRegion.getSecond() > 2 * avgRegionSize) {
LOG.debug("Table " + table + ", largest region "
+ largestRegion.getFirst().getRegionNameAsString() + " has size "
+ largestRegion.getSecond() + ", more than 2 times than avg size, splitting");
return new SplitNormalizationPlan(largestRegion.getFirst(), null);
}
int candidateIdx = 0;
// look for two successive entries whose indices are adjacent
while (candidateIdx < tableRegions.size()-1) {
if (Math.abs(regionsWithSize.get(candidateIdx).getThird() -
regionsWithSize.get(candidateIdx + 1).getThird()) == 1) {
break;
while (candidateIdx < tableRegions.size()) {
HRegionInfo hri = tableRegions.get(candidateIdx);
long regionSize = getRegionSize(hri);
// if the region is > 2 times larger than average, we split it, split
// is more high priority normalization action than merge.
if (types.contains(PlanType.SPLIT) && regionSize > 2 * avgRegionSize) {
LOG.debug("Table " + table + ", large region " + hri.getRegionNameAsString() + " has size "
+ regionSize + ", more than twice avg size, splitting");
plans.add(new SplitNormalizationPlan(hri, null));
} else {
if (candidateIdx == tableRegions.size()-1) {
break;
}
HRegionInfo hri2 = tableRegions.get(candidateIdx+1);
long regionSize2 = getRegionSize(hri2);
if (types.contains(PlanType.MERGE) && regionSize + regionSize2 < avgRegionSize) {
LOG.debug("Table " + table + ", small region size: " + regionSize
+ " plus its neighbor size: " + regionSize2
+ ", less than the avg size " + avgRegionSize + ", merging them");
plans.add(new MergeNormalizationPlan(hri, hri2));
candidateIdx++;
}
}
candidateIdx++;
}
if (candidateIdx == tableRegions.size()-1) {
LOG.debug("No neighboring regions found for table: " + table);
return EmptyNormalizationPlan.getInstance();
if (plans.isEmpty()) {
LOG.debug("No normalization needed, regions look good for table: " + table);
return null;
}
Triple<HRegionInfo, Long, Integer> candidateRegion = regionsWithSize.get(candidateIdx);
Triple<HRegionInfo, Long, Integer> candidateRegion2 = regionsWithSize.get(candidateIdx+1);
if (types.contains(PlanType.MERGE) &&
candidateRegion.getSecond() + candidateRegion2.getSecond() < avgRegionSize) {
LOG.debug("Table " + table + ", smallest region size: " + candidateRegion.getSecond()
+ " and its smallest neighbor size: " + candidateRegion2.getSecond()
+ ", less than the avg size, merging them");
return new MergeNormalizationPlan(candidateRegion.getFirst(),
candidateRegion2.getFirst());
}
LOG.debug("No normalization needed, regions look good for table: " + table);
return EmptyNormalizationPlan.getInstance();
Collections.sort(plans, planComparator);
return plans;
}
private long getRegionSize(HRegionInfo hri) {

27
hbase-server/src/test/java/org/apache/hadoop/hbase/master/normalizer/TestSimpleRegionNormalizer.java
View File

@ -83,8 +83,8 @@ public class TestSimpleRegionNormalizer {
Map<byte[], Integer> regionSizes = new HashMap<>();
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plan instanceof EmptyNormalizationPlan);
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plans == null);
}
@Test
@ -102,8 +102,8 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri2.getRegionName(), 15);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue((plan instanceof EmptyNormalizationPlan));
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plans == null);
}
@Test
@ -129,8 +129,8 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri4.getRegionName(), 10);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plan instanceof EmptyNormalizationPlan);
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plans == null);
}
@Test
@ -165,15 +165,16 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri5.getRegionName(), 16);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable,
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable,
mergeDesired ? bothTypes : splitType);
if (mergeDesired) {
NormalizationPlan plan = plans.get(0);
assertTrue(plan instanceof MergeNormalizationPlan);
assertEquals(hri2, ((MergeNormalizationPlan) plan).getFirstRegion());
assertEquals(hri3, ((MergeNormalizationPlan) plan).getSecondRegion());
} else {
assertTrue(plan instanceof EmptyNormalizationPlan);
assertTrue(plans == null);
}
}
@ -209,7 +210,8 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri6.getRegionName(), 2700);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
NormalizationPlan plan = plans.get(0);
assertTrue(plan instanceof MergeNormalizationPlan);
assertEquals(hri5, ((MergeNormalizationPlan) plan).getFirstRegion());
@ -243,9 +245,9 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri5.getRegionName(), 5);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
assertTrue(plan instanceof EmptyNormalizationPlan);
assertTrue(plans == null);
}
@Test
@ -271,7 +273,8 @@ public class TestSimpleRegionNormalizer {
regionSizes.put(hri4.getRegionName(), 30);
setupMocksForNormalizer(regionSizes, hris);
NormalizationPlan plan = normalizer.computePlanForTable(testTable, bothTypes);
List<NormalizationPlan> plans = normalizer.computePlanForTable(testTable, bothTypes);
NormalizationPlan plan = plans.get(0);
assertTrue(plan instanceof SplitNormalizationPlan);
assertEquals(hri4, ((SplitNormalizationPlan) plan).getRegionInfo());

16
hbase-server/src/test/java/org/apache/hadoop/hbase/master/normalizer/TestSimpleRegionNormalizerOnCluster.java
View File

@ -155,11 +155,19 @@ public class TestSimpleRegionNormalizerOnCluster {
} while (skippedSplitcnt == 0L);
assert(skippedSplitcnt > 0);
} else {
while (MetaTableAccessor.getRegionCount(TEST_UTIL.getConnection(), TABLENAME) < 6) {
LOG.info("Waiting for normalization split to complete");
Thread.sleep(100);
while (true) {
List<HRegion> regions = TEST_UTIL.getHBaseCluster().getRegions(TABLENAME);
int cnt = 0;
for (HRegion region : regions) {
String regionName = region.getRegionInfo().getRegionNameAsString();
if (regionName.startsWith("testRegionNormalizationSplitOnCluster,zzzzz")) {
cnt++;
}
}
if (cnt >= 2) {
break;
}
}
assertEquals(6, MetaTableAccessor.getRegionCount(TEST_UTIL.getConnection(), TABLENAME));
}
admin.disableTable(TABLENAME);