HBASE-23350 Make compaction files cacheonWrite configurable based on threshold

Signed-off-by: ramkrish86 <ramkrishna@apache.org>
This commit is contained in:
abhinaba.sarkar 2020-01-31 11:35:11 +05:30 committed by ramkrish86
parent 76247aa21f
commit bf924ccdaa
8 changed files with 143 additions and 22 deletions

View File

@ -87,6 +87,13 @@ public class CacheConfig {
public static final String CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY =
"hbase.rs.cachecompactedblocksonwrite";
/**
* Configuration key to determine total size in bytes of compacted files beyond which we do not
* cache blocks on compaction
*/
public static final String CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD_KEY =
"hbase.rs.cachecompactedblocksonwrite.threshold";
public static final String DROP_BEHIND_CACHE_COMPACTION_KEY =
"hbase.hfile.drop.behind.compaction";
@ -101,6 +108,7 @@ public class CacheConfig {
public static final boolean DEFAULT_PREFETCH_ON_OPEN = false;
public static final boolean DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE = false;
public static final boolean DROP_BEHIND_CACHE_COMPACTION_DEFAULT = true;
public static final long DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD = Long.MAX_VALUE;
/**
* Whether blocks should be cached on read (default is on if there is a
@ -136,6 +144,11 @@ public class CacheConfig {
*/
private final boolean cacheCompactedDataOnWrite;
/**
* Determine threshold beyond which we do not cache blocks on compaction
*/
private long cacheCompactedDataOnWriteThreshold;
private final boolean dropBehindCompaction;
// Local reference to the block cache
@ -188,6 +201,7 @@ public class CacheConfig {
(family == null ? false : family.isPrefetchBlocksOnOpen());
this.cacheCompactedDataOnWrite = conf.getBoolean(CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY,
DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE);
this.cacheCompactedDataOnWriteThreshold = getCacheCompactedBlocksOnWriteThreshold(conf);
this.blockCache = blockCache;
this.byteBuffAllocator = byteBuffAllocator;
LOG.info("Created cacheConfig: " + this + (family == null ? "" : " for family " + family) +
@ -208,6 +222,7 @@ public class CacheConfig {
this.cacheDataCompressed = cacheConf.cacheDataCompressed;
this.prefetchOnOpen = cacheConf.prefetchOnOpen;
this.cacheCompactedDataOnWrite = cacheConf.cacheCompactedDataOnWrite;
this.cacheCompactedDataOnWriteThreshold = cacheConf.cacheCompactedDataOnWriteThreshold;
this.dropBehindCompaction = cacheConf.dropBehindCompaction;
this.blockCache = cacheConf.blockCache;
this.byteBuffAllocator = cacheConf.byteBuffAllocator;
@ -275,7 +290,6 @@ public class CacheConfig {
this.cacheDataOnWrite = cacheDataOnWrite;
}
/**
* Enable cache on write including:
* cacheDataOnWrite
@ -288,7 +302,6 @@ public class CacheConfig {
this.cacheBloomsOnWrite = true;
}
/**
* @return true if index blocks should be written to the cache when an HFile
* is written, false if not
@ -356,6 +369,12 @@ public class CacheConfig {
return this.cacheCompactedDataOnWrite;
}
/**
* @return total file size in bytes threshold for caching while writing during compaction
*/
public long getCacheCompactedBlocksOnWriteThreshold() {
return this.cacheCompactedDataOnWriteThreshold;
}
/**
* Return true if we may find this type of block in block cache.
* <p>
@ -412,6 +431,21 @@ public class CacheConfig {
return this.byteBuffAllocator;
}
private long getCacheCompactedBlocksOnWriteThreshold(Configuration conf) {
long cacheCompactedBlocksOnWriteThreshold = conf
.getLong(CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD_KEY,
DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD);
if (cacheCompactedBlocksOnWriteThreshold < 0) {
LOG.warn(
"cacheCompactedBlocksOnWriteThreshold value : {} is less than 0, resetting it to: {}",
cacheCompactedBlocksOnWriteThreshold, DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD);
cacheCompactedBlocksOnWriteThreshold = DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD;
}
return cacheCompactedBlocksOnWriteThreshold;
}
@Override
public String toString() {
return "cacheDataOnRead=" + shouldCacheDataOnRead() + ", cacheDataOnWrite="

View File

@ -1110,6 +1110,13 @@ public class HStore implements Store, HeapSize, StoreConfigInformation,
return sf;
}
public StoreFileWriter createWriterInTmp(long maxKeyCount, Compression.Algorithm compression,
boolean isCompaction, boolean includeMVCCReadpoint, boolean includesTag,
boolean shouldDropBehind) throws IOException {
return createWriterInTmp(maxKeyCount, compression, isCompaction, includeMVCCReadpoint,
includesTag, shouldDropBehind, -1);
}
/**
* @param compression Compression algorithm to use
* @param isCompaction whether we are creating a new file in a compaction
@ -1121,17 +1128,19 @@ public class HStore implements Store, HeapSize, StoreConfigInformation,
// compaction
public StoreFileWriter createWriterInTmp(long maxKeyCount, Compression.Algorithm compression,
boolean isCompaction, boolean includeMVCCReadpoint, boolean includesTag,
boolean shouldDropBehind) throws IOException {
final CacheConfig writerCacheConf;
boolean shouldDropBehind, long totalCompactedFilesSize) throws IOException {
// creating new cache config for each new writer
final CacheConfig writerCacheConf = new CacheConfig(cacheConf);
if (isCompaction) {
// Don't cache data on write on compactions, unless specifically configured to do so
writerCacheConf = new CacheConfig(cacheConf);
// Cache only when total file size remains lower than configured threshold
final boolean cacheCompactedBlocksOnWrite =
cacheConf.shouldCacheCompactedBlocksOnWrite();
// if data blocks are to be cached on write
// during compaction, we should forcefully
// cache index and bloom blocks as well
if (cacheCompactedBlocksOnWrite) {
if (cacheCompactedBlocksOnWrite && totalCompactedFilesSize <= cacheConf
.getCacheCompactedBlocksOnWriteThreshold()) {
writerCacheConf.enableCacheOnWrite();
if (!cacheOnWriteLogged) {
LOG.info("For Store {} , cacheCompactedBlocksOnWrite is true, hence enabled " +
@ -1141,9 +1150,16 @@ public class HStore implements Store, HeapSize, StoreConfigInformation,
}
} else {
writerCacheConf.setCacheDataOnWrite(false);
if (totalCompactedFilesSize > cacheConf.getCacheCompactedBlocksOnWriteThreshold()) {
// checking condition once again for logging
LOG.debug(
"For Store {}, setting cacheCompactedBlocksOnWrite as false as total size of compacted "
+ "files - {}, is greater than cacheCompactedBlocksOnWriteThreshold - {}",
getColumnFamilyName(), totalCompactedFilesSize,
cacheConf.getCacheCompactedBlocksOnWriteThreshold());
}
}
} else {
writerCacheConf = cacheConf;
final boolean shouldCacheDataOnWrite = cacheConf.shouldCacheDataOnWrite();
if (shouldCacheDataOnWrite) {
writerCacheConf.enableCacheOnWrite();

View File

@ -130,6 +130,8 @@ public abstract class Compactor<T extends CellSink> {
public int maxTagsLength = 0;
/** Min SeqId to keep during a major compaction **/
public long minSeqIdToKeep = 0;
/** Total size of the compacted files **/
private long totalCompactedFilesSize = 0;
}
/**
@ -166,6 +168,10 @@ public abstract class Compactor<T extends CellSink> {
fd.maxKeyCount += keyCount;
// calculate the latest MVCC readpoint in any of the involved store files
Map<byte[], byte[]> fileInfo = r.loadFileInfo();
// calculate the total size of the compacted files
fd.totalCompactedFilesSize += r.length();
byte[] tmp = null;
// Get and set the real MVCCReadpoint for bulk loaded files, which is the
// SeqId number.
@ -260,8 +266,9 @@ public abstract class Compactor<T extends CellSink> {
throws IOException {
// When all MVCC readpoints are 0, don't write them.
// See HBASE-8166, HBASE-12600, and HBASE-13389.
return store.createWriterInTmp(fd.maxKeyCount, this.compactionCompression, true,
fd.maxMVCCReadpoint > 0, fd.maxTagsLength > 0, shouldDropBehind);
return store
.createWriterInTmp(fd.maxKeyCount, this.compactionCompression, true, fd.maxMVCCReadpoint > 0,
fd.maxTagsLength > 0, shouldDropBehind, fd.totalCompactedFilesSize);
}
private ScanInfo preCompactScannerOpen(CompactionRequestImpl request, ScanType scanType,

View File

@ -5424,8 +5424,9 @@ public class TestFromClientSide {
System.out.println("Flushing cache");
region.flush(true);
// + 1 for Index Block
assertEquals(++expectedBlockCount + 1, cache.getBlockCount());
// + 1 for Index Block, +1 for data block
expectedBlockCount += 2;
assertEquals(expectedBlockCount, cache.getBlockCount());
assertEquals(expectedBlockHits, cache.getStats().getHitCount());
assertEquals(expectedBlockMiss, cache.getStats().getMissCount());
// compact, net minus two blocks, two hits, no misses
@ -5436,7 +5437,8 @@ public class TestFromClientSide {
store.closeAndArchiveCompactedFiles();
waitForStoreFileCount(store, 1, 10000); // wait 10 seconds max
assertEquals(1, store.getStorefilesCount());
expectedBlockCount -= 2; // evicted two blocks, cached none
// evicted two data blocks and two index blocks and compaction does not cache new blocks
expectedBlockCount = 0;
assertEquals(expectedBlockCount, cache.getBlockCount());
expectedBlockHits += 2;
assertEquals(expectedBlockMiss, cache.getStats().getMissCount());

View File

@ -127,6 +127,10 @@ public class TestCacheOnWrite {
BlockType.DATA
);
// All test cases are supposed to generate files for compaction within this range
private static final long CACHE_COMPACTION_LOW_THRESHOLD = 10L;
private static final long CACHE_COMPACTION_HIGH_THRESHOLD = 1 * 1024 * 1024 * 1024L;
/** The number of valid key types possible in a store file */
private static final int NUM_VALID_KEY_TYPES =
KeyValue.Type.values().length - 2;
@ -424,15 +428,31 @@ public class TestCacheOnWrite {
}
private void testCachingDataBlocksDuringCompactionInternals(boolean useTags,
boolean cacheBlocksOnCompaction) throws IOException, InterruptedException {
boolean cacheBlocksOnCompaction, long cacheBlocksOnCompactionThreshold)
throws IOException, InterruptedException {
// create a localConf
boolean localValue = conf.getBoolean(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY,
false);
boolean localValue = conf.getBoolean(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY, false);
long localCacheCompactedBlocksThreshold = conf
.getLong(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD_KEY,
CacheConfig.DEFAULT_CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD);
boolean localCacheBloomBlocksValue = conf
.getBoolean(CacheConfig.CACHE_BLOOM_BLOCKS_ON_WRITE_KEY,
CacheConfig.DEFAULT_CACHE_BLOOMS_ON_WRITE);
boolean localCacheIndexBlocksValue = conf
.getBoolean(CacheConfig.CACHE_INDEX_BLOCKS_ON_WRITE_KEY,
CacheConfig.DEFAULT_CACHE_INDEXES_ON_WRITE);
try {
// Set the conf if testing caching compacted blocks on write
conf.setBoolean(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY,
cacheBlocksOnCompaction);
// set size threshold if testing compaction size threshold
if (cacheBlocksOnCompactionThreshold > 0) {
conf.setLong(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD_KEY,
cacheBlocksOnCompactionThreshold);
}
// TODO: need to change this test if we add a cache size threshold for
// compactions, or if we implement some other kind of intelligent logic for
// deciding what blocks to cache-on-write on compaction.
@ -467,7 +487,9 @@ public class TestCacheOnWrite {
HConstants.LATEST_TIMESTAMP, Bytes.toBytes(valueStr), tags);
p.add(kv);
} else {
p.addColumn(cfBytes, Bytes.toBytes(qualStr), ts++, Bytes.toBytes(valueStr));
KeyValue kv = new KeyValue(Bytes.toBytes(rowStr), cfBytes, Bytes.toBytes(qualStr),
ts++, Bytes.toBytes(valueStr));
p.add(kv);
}
}
}
@ -507,11 +529,33 @@ public class TestCacheOnWrite {
"\ncacheBlocksOnCompaction: "
+ cacheBlocksOnCompaction + "\n";
assertEquals(assertErrorMessage, cacheOnCompactAndNonBucketCache, dataBlockCached);
if (cacheOnCompactAndNonBucketCache && cacheBlocksOnCompactionThreshold > 0) {
if (cacheBlocksOnCompactionThreshold == CACHE_COMPACTION_HIGH_THRESHOLD) {
assertTrue(assertErrorMessage, dataBlockCached);
assertTrue(assertErrorMessage, bloomBlockCached);
assertTrue(assertErrorMessage, indexBlockCached);
} else {
assertFalse(assertErrorMessage, dataBlockCached);
if (cacheOnCompactAndNonBucketCache) {
assertTrue(assertErrorMessage, bloomBlockCached);
assertTrue(assertErrorMessage, indexBlockCached);
if (localCacheBloomBlocksValue) {
assertTrue(assertErrorMessage, bloomBlockCached);
} else {
assertFalse(assertErrorMessage, bloomBlockCached);
}
if (localCacheIndexBlocksValue) {
assertTrue(assertErrorMessage, indexBlockCached);
} else {
assertFalse(assertErrorMessage, indexBlockCached);
}
}
} else {
assertEquals(assertErrorMessage, cacheOnCompactAndNonBucketCache, dataBlockCached);
if (cacheOnCompactAndNonBucketCache) {
assertTrue(assertErrorMessage, bloomBlockCached);
assertTrue(assertErrorMessage, indexBlockCached);
}
}
@ -519,6 +563,10 @@ public class TestCacheOnWrite {
} finally {
// reset back
conf.setBoolean(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_KEY, localValue);
conf.setLong(CacheConfig.CACHE_COMPACTED_BLOCKS_ON_WRITE_THRESHOLD_KEY,
localCacheCompactedBlocksThreshold);
conf.setBoolean(CacheConfig.CACHE_BLOOM_BLOCKS_ON_WRITE_KEY, localCacheBloomBlocksValue);
conf.setBoolean(CacheConfig.CACHE_INDEX_BLOCKS_ON_WRITE_KEY, localCacheIndexBlocksValue);
}
}
@ -530,8 +578,15 @@ public class TestCacheOnWrite {
@Test
public void testCachingDataBlocksDuringCompaction() throws IOException, InterruptedException {
testCachingDataBlocksDuringCompactionInternals(false, false);
testCachingDataBlocksDuringCompactionInternals(true, true);
testCachingDataBlocksDuringCompactionInternals(false, false, -1);
testCachingDataBlocksDuringCompactionInternals(true, true, -1);
}
@Test
public void testCachingDataBlocksThresholdDuringCompaction()
throws IOException, InterruptedException {
testCachingDataBlocksDuringCompactionInternals(false, true, CACHE_COMPACTION_HIGH_THRESHOLD);
testCachingDataBlocksDuringCompactionInternals(false, true, CACHE_COMPACTION_LOW_THRESHOLD);
}
}

View File

@ -108,6 +108,8 @@ public class TestDateTieredCompactor {
when(store.getRegionInfo()).thenReturn(new HRegionInfo(TABLE_NAME));
when(store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean())).thenAnswer(writers);
when(store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean(), anyLong())).thenAnswer(writers);
when(store.getComparator()).thenReturn(CellComparatorImpl.COMPARATOR);
OptionalLong maxSequenceId = StoreUtils.getMaxSequenceIdInList(storefiles);
when(store.getMaxSequenceId()).thenReturn(maxSequenceId);

View File

@ -785,6 +785,9 @@ public class TestStripeCompactionPolicy {
when(
store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean())).thenAnswer(writers);
when(
store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean(), anyLong())).thenAnswer(writers);
Configuration conf = HBaseConfiguration.create();
conf.setBoolean("hbase.regionserver.compaction.private.readers", usePrivateReaders);

View File

@ -208,6 +208,8 @@ public class TestStripeCompactor {
when(store.getRegionInfo()).thenReturn(new HRegionInfo(TABLE_NAME));
when(store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean())).thenAnswer(writers);
when(store.createWriterInTmp(anyLong(), any(), anyBoolean(),
anyBoolean(), anyBoolean(), anyBoolean(), anyLong())).thenAnswer(writers);
when(store.getComparator()).thenReturn(CellComparatorImpl.COMPARATOR);
return new StripeCompactor(conf, store) {