org.apache.cassandra.service.CacheService Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of cassandra-all Show documentation
Show all versions of cassandra-all Show documentation
The Apache Cassandra Project develops a highly scalable second-generation distributed database, bringing together Dynamo's fully distributed design and Bigtable's ColumnFamily-based data model.
/*
* Licensed to the Apache Software Foundation (ASF) under one
* or more contributor license agreements. See the NOTICE file
* distributed with this work for additional information
* regarding copyright ownership. The ASF licenses this file
* to you under the Apache License, Version 2.0 (the
* "License"); you may not use this file except in compliance
* with the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package org.apache.cassandra.service;
import java.io.IOException;
import java.nio.ByteBuffer;
import java.util.ArrayList;
import java.util.HashMap;
import java.util.Iterator;
import java.util.List;
import java.util.Map;
import java.util.concurrent.Callable;
import java.util.concurrent.ConcurrentHashMap;
import java.util.concurrent.ExecutionException;
import java.util.concurrent.Future;
import com.google.common.util.concurrent.Futures;
import org.apache.cassandra.db.lifecycle.SSTableSet;
import org.apache.cassandra.io.sstable.format.SSTableReader;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.apache.cassandra.cache.*;
import org.apache.cassandra.cache.AutoSavingCache.CacheSerializer;
import org.apache.cassandra.concurrent.Stage;
import org.apache.cassandra.concurrent.StageManager;
import org.apache.cassandra.config.CFMetaData;
import org.apache.cassandra.config.ColumnDefinition;
import org.apache.cassandra.config.DatabaseDescriptor;
import org.apache.cassandra.config.Schema;
import org.apache.cassandra.db.*;
import org.apache.cassandra.db.rows.*;
import org.apache.cassandra.db.filter.*;
import org.apache.cassandra.db.partitions.CachedBTreePartition;
import org.apache.cassandra.db.partitions.CachedPartition;
import org.apache.cassandra.db.context.CounterContext;
import org.apache.cassandra.io.util.DataInputPlus;
import org.apache.cassandra.io.util.DataOutputPlus;
import org.apache.cassandra.io.sstable.format.big.BigFormat;
import org.apache.cassandra.utils.ByteBufferUtil;
import org.apache.cassandra.utils.FBUtilities;
import org.apache.cassandra.utils.MBeanWrapper;
import org.apache.cassandra.utils.Pair;
import org.apache.cassandra.utils.concurrent.OpOrder;
public class CacheService implements CacheServiceMBean
{
private static final Logger logger = LoggerFactory.getLogger(CacheService.class);
public static final String MBEAN_NAME = "org.apache.cassandra.db:type=Caches";
public static enum CacheType
{
KEY_CACHE("KeyCache"),
ROW_CACHE("RowCache"),
COUNTER_CACHE("CounterCache");
private final String name;
private CacheType(String typeName)
{
name = typeName;
}
public String toString()
{
return name;
}
}
public final static CacheService instance = new CacheService();
public final AutoSavingCache keyCache;
public final AutoSavingCache rowCache;
public final AutoSavingCache counterCache;
private CacheService()
{
MBeanWrapper.instance.registerMBean(this, MBEAN_NAME);
keyCache = initKeyCache();
rowCache = initRowCache();
counterCache = initCounterCache();
}
/**
* @return auto saving cache object
*/
private AutoSavingCache initKeyCache()
{
logger.info("Initializing key cache with capacity of {} MBs.", DatabaseDescriptor.getKeyCacheSizeInMB());
long keyCacheInMemoryCapacity = DatabaseDescriptor.getKeyCacheSizeInMB() * 1024 * 1024;
// as values are constant size we can use singleton weigher
// where 48 = 40 bytes (average size of the key) + 8 bytes (size of value)
ICache kc;
kc = ConcurrentLinkedHashCache.create(keyCacheInMemoryCapacity);
AutoSavingCache keyCache = new AutoSavingCache<>(kc, CacheType.KEY_CACHE, new KeyCacheSerializer());
int keyCacheKeysToSave = DatabaseDescriptor.getKeyCacheKeysToSave();
keyCache.scheduleSaving(DatabaseDescriptor.getKeyCacheSavePeriod(), keyCacheKeysToSave);
return keyCache;
}
/**
* @return initialized row cache
*/
private AutoSavingCache initRowCache()
{
logger.info("Initializing row cache with capacity of {} MBs", DatabaseDescriptor.getRowCacheSizeInMB());
CacheProvider cacheProvider;
String cacheProviderClassName = DatabaseDescriptor.getRowCacheSizeInMB() > 0
? DatabaseDescriptor.getRowCacheClassName() : "org.apache.cassandra.cache.NopCacheProvider";
try
{
Class> cacheProviderClass =
(Class>) Class.forName(cacheProviderClassName);
cacheProvider = cacheProviderClass.newInstance();
}
catch (Exception e)
{
throw new RuntimeException("Cannot find configured row cache provider class " + DatabaseDescriptor.getRowCacheClassName());
}
// cache object
ICache rc = cacheProvider.create();
AutoSavingCache rowCache = new AutoSavingCache<>(rc, CacheType.ROW_CACHE, new RowCacheSerializer());
int rowCacheKeysToSave = DatabaseDescriptor.getRowCacheKeysToSave();
rowCache.scheduleSaving(DatabaseDescriptor.getRowCacheSavePeriod(), rowCacheKeysToSave);
return rowCache;
}
private AutoSavingCache initCounterCache()
{
logger.info("Initializing counter cache with capacity of {} MBs", DatabaseDescriptor.getCounterCacheSizeInMB());
long capacity = DatabaseDescriptor.getCounterCacheSizeInMB() * 1024 * 1024;
AutoSavingCache cache =
new AutoSavingCache<>(ConcurrentLinkedHashCache.create(capacity),
CacheType.COUNTER_CACHE,
new CounterCacheSerializer());
int keysToSave = DatabaseDescriptor.getCounterCacheKeysToSave();
logger.info("Scheduling counter cache save to every {} seconds (going to save {} keys).",
DatabaseDescriptor.getCounterCacheSavePeriod(),
keysToSave == Integer.MAX_VALUE ? "all" : keysToSave);
cache.scheduleSaving(DatabaseDescriptor.getCounterCacheSavePeriod(), keysToSave);
return cache;
}
public int getRowCacheSavePeriodInSeconds()
{
return DatabaseDescriptor.getRowCacheSavePeriod();
}
public void setRowCacheSavePeriodInSeconds(int seconds)
{
if (seconds < 0)
throw new RuntimeException("RowCacheSavePeriodInSeconds must be non-negative.");
DatabaseDescriptor.setRowCacheSavePeriod(seconds);
rowCache.scheduleSaving(seconds, DatabaseDescriptor.getRowCacheKeysToSave());
}
public int getKeyCacheSavePeriodInSeconds()
{
return DatabaseDescriptor.getKeyCacheSavePeriod();
}
public void setKeyCacheSavePeriodInSeconds(int seconds)
{
if (seconds < 0)
throw new RuntimeException("KeyCacheSavePeriodInSeconds must be non-negative.");
DatabaseDescriptor.setKeyCacheSavePeriod(seconds);
keyCache.scheduleSaving(seconds, DatabaseDescriptor.getKeyCacheKeysToSave());
}
public int getCounterCacheSavePeriodInSeconds()
{
return DatabaseDescriptor.getCounterCacheSavePeriod();
}
public void setCounterCacheSavePeriodInSeconds(int seconds)
{
if (seconds < 0)
throw new RuntimeException("CounterCacheSavePeriodInSeconds must be non-negative.");
DatabaseDescriptor.setCounterCacheSavePeriod(seconds);
counterCache.scheduleSaving(seconds, DatabaseDescriptor.getCounterCacheKeysToSave());
}
public int getRowCacheKeysToSave()
{
return DatabaseDescriptor.getRowCacheKeysToSave();
}
public void setRowCacheKeysToSave(int count)
{
if (count < 0)
throw new RuntimeException("RowCacheKeysToSave must be non-negative.");
DatabaseDescriptor.setRowCacheKeysToSave(count);
rowCache.scheduleSaving(getRowCacheSavePeriodInSeconds(), count);
}
public int getKeyCacheKeysToSave()
{
return DatabaseDescriptor.getKeyCacheKeysToSave();
}
public void setKeyCacheKeysToSave(int count)
{
if (count < 0)
throw new RuntimeException("KeyCacheKeysToSave must be non-negative.");
DatabaseDescriptor.setKeyCacheKeysToSave(count);
keyCache.scheduleSaving(getKeyCacheSavePeriodInSeconds(), count);
}
public int getCounterCacheKeysToSave()
{
return DatabaseDescriptor.getCounterCacheKeysToSave();
}
public void setCounterCacheKeysToSave(int count)
{
if (count < 0)
throw new RuntimeException("CounterCacheKeysToSave must be non-negative.");
DatabaseDescriptor.setCounterCacheKeysToSave(count);
counterCache.scheduleSaving(getCounterCacheSavePeriodInSeconds(), count);
}
public void invalidateKeyCache()
{
keyCache.clear();
}
public void invalidateKeyCacheForCf(Pair ksAndCFName)
{
Iterator keyCacheIterator = keyCache.keyIterator();
while (keyCacheIterator.hasNext())
{
KeyCacheKey key = keyCacheIterator.next();
if (key.ksAndCFName.equals(ksAndCFName))
keyCacheIterator.remove();
}
}
public void invalidateRowCache()
{
rowCache.clear();
}
public void invalidateRowCacheForCf(Pair ksAndCFName)
{
Iterator rowCacheIterator = rowCache.keyIterator();
while (rowCacheIterator.hasNext())
{
RowCacheKey rowCacheKey = rowCacheIterator.next();
if (rowCacheKey.ksAndCFName.equals(ksAndCFName))
rowCacheIterator.remove();
}
}
public void invalidateCounterCacheForCf(Pair ksAndCFName)
{
Iterator counterCacheIterator = counterCache.keyIterator();
while (counterCacheIterator.hasNext())
{
CounterCacheKey counterCacheKey = counterCacheIterator.next();
if (counterCacheKey.ksAndCFName.equals(ksAndCFName))
counterCacheIterator.remove();
}
}
public void invalidateCounterCache()
{
counterCache.clear();
}
public void setRowCacheCapacityInMB(long capacity)
{
if (capacity < 0)
throw new RuntimeException("capacity should not be negative.");
rowCache.setCapacity(capacity * 1024 * 1024);
}
public void setKeyCacheCapacityInMB(long capacity)
{
if (capacity < 0)
throw new RuntimeException("capacity should not be negative.");
keyCache.setCapacity(capacity * 1024 * 1024);
}
public void setCounterCacheCapacityInMB(long capacity)
{
if (capacity < 0)
throw new RuntimeException("capacity should not be negative.");
counterCache.setCapacity(capacity * 1024 * 1024);
}
public void saveCaches() throws ExecutionException, InterruptedException
{
List> futures = new ArrayList<>(3);
logger.debug("submitting cache saves");
futures.add(keyCache.submitWrite(DatabaseDescriptor.getKeyCacheKeysToSave()));
futures.add(rowCache.submitWrite(DatabaseDescriptor.getRowCacheKeysToSave()));
futures.add(counterCache.submitWrite(DatabaseDescriptor.getCounterCacheKeysToSave()));
FBUtilities.waitOnFutures(futures);
logger.debug("cache saves completed");
}
public static class CounterCacheSerializer implements CacheSerializer
{
public void serialize(CounterCacheKey key, DataOutputPlus out, ColumnFamilyStore cfs) throws IOException
{
assert(cfs.metadata.isCounter());
out.write(cfs.metadata.ksAndCFBytes);
ByteBufferUtil.writeWithLength(key.partitionKey, out);
ByteBufferUtil.writeWithLength(key.cellName, out);
}
public Future> deserialize(DataInputPlus in, final ColumnFamilyStore cfs) throws IOException
{
//Keyspace and CF name are deserialized by AutoSaving cache and used to fetch the CFS provided as a
//parameter so they aren't deserialized here, even though they are serialized by this serializer
final ByteBuffer partitionKey = ByteBufferUtil.readWithLength(in);
final ByteBuffer cellName = ByteBufferUtil.readWithLength(in);
if (cfs == null || !cfs.metadata.isCounter() || !cfs.isCounterCacheEnabled())
return null;
assert(cfs.metadata.isCounter());
return StageManager.getStage(Stage.READ).submit(new Callable>()
{
public Pair call() throws Exception
{
DecoratedKey key = cfs.decorateKey(partitionKey);
LegacyLayout.LegacyCellName name = LegacyLayout.decodeCellName(cfs.metadata, cellName);
ColumnDefinition column = name.column;
CellPath path = name.collectionElement == null ? null : CellPath.create(name.collectionElement);
int nowInSec = FBUtilities.nowInSeconds();
ColumnFilter.Builder builder = ColumnFilter.selectionBuilder();
if (path == null)
builder.add(column);
else
builder.select(column, path);
ClusteringIndexFilter filter = new ClusteringIndexNamesFilter(FBUtilities.singleton(name.clustering, cfs.metadata.comparator), false);
SinglePartitionReadCommand cmd = SinglePartitionReadCommand.create(cfs.metadata, nowInSec, key, builder.build(), filter);
try (OpOrder.Group op = cfs.readOrdering.start(); RowIterator iter = UnfilteredRowIterators.filter(cmd.queryMemtableAndDisk(cfs, op), nowInSec))
{
Cell cell;
if (column.isStatic())
{
cell = iter.staticRow().getCell(column);
}
else
{
if (!iter.hasNext())
return null;
cell = iter.next().getCell(column);
}
if (cell == null)
return null;
ClockAndCount clockAndCount = CounterContext.instance().getLocalClockAndCount(cell.value());
return Pair.create(CounterCacheKey.create(cfs.metadata.ksAndCFName, partitionKey, name.clustering, column, path), clockAndCount);
}
}
});
}
}
public static class RowCacheSerializer implements CacheSerializer
{
public void serialize(RowCacheKey key, DataOutputPlus out, ColumnFamilyStore cfs) throws IOException
{
assert(!cfs.isIndex());//Shouldn't have row cache entries for indexes
out.write(cfs.metadata.ksAndCFBytes);
ByteBufferUtil.writeWithLength(key.key, out);
}
public Future> deserialize(DataInputPlus in, final ColumnFamilyStore cfs) throws IOException
{
//Keyspace and CF name are deserialized by AutoSaving cache and used to fetch the CFS provided as a
//parameter so they aren't deserialized here, even though they are serialized by this serializer
final ByteBuffer buffer = ByteBufferUtil.readWithLength(in);
final int rowsToCache = cfs.metadata.params.caching.rowsPerPartitionToCache();
if (cfs == null || !cfs.isRowCacheEnabled())
return null;
assert(!cfs.isIndex());//Shouldn't have row cache entries for indexes
return StageManager.getStage(Stage.READ).submit(new Callable>()
{
public Pair call() throws Exception
{
DecoratedKey key = cfs.decorateKey(buffer);
int nowInSec = FBUtilities.nowInSeconds();
try (OpOrder.Group op = cfs.readOrdering.start(); UnfilteredRowIterator iter = SinglePartitionReadCommand.fullPartitionRead(cfs.metadata, nowInSec, key).queryMemtableAndDisk(cfs, op))
{
CachedPartition toCache = CachedBTreePartition.create(DataLimits.cqlLimits(rowsToCache).filter(iter, nowInSec, true), nowInSec);
return Pair.create(new RowCacheKey(cfs.metadata.ksAndCFName, key), (IRowCacheEntry)toCache);
}
}
});
}
}
public static class KeyCacheSerializer implements CacheSerializer
{
// For column families with many SSTables the linear nature of getSSTables slowed down KeyCache loading
// by orders of magnitude. So we cache the sstables once and rely on cleanupAfterDeserialize to cleanup any
// cached state we may have accumulated during the load.
Map, Map> cachedSSTableReaders = new ConcurrentHashMap<>();
public void serialize(KeyCacheKey key, DataOutputPlus out, ColumnFamilyStore cfs) throws IOException
{
//Don't serialize old format entries since we didn't bother to implement serialization of both for simplicity
//https://issues.apache.org/jira/browse/CASSANDRA-10778
if (!key.desc.version.storeRows()) return;
RowIndexEntry entry = CacheService.instance.keyCache.getInternal(key);
if (entry == null)
return;
out.write(cfs.metadata.ksAndCFBytes);
ByteBufferUtil.writeWithLength(key.key, out);
out.writeInt(key.desc.generation);
out.writeBoolean(true);
key.desc.getFormat().getIndexSerializer(cfs.metadata, key.desc.version, SerializationHeader.forKeyCache(cfs.metadata)).serialize(entry, out);
}
public Future> deserialize(DataInputPlus input, ColumnFamilyStore cfs) throws IOException
{
boolean skipEntry = cfs == null || !cfs.isKeyCacheEnabled();
//Keyspace and CF name are deserialized by AutoSaving cache and used to fetch the CFS provided as a
//parameter so they aren't deserialized here, even though they are serialized by this serializer
int keyLength = input.readInt();
if (keyLength > FBUtilities.MAX_UNSIGNED_SHORT)
{
throw new IOException(String.format("Corrupted key cache. Key length of %d is longer than maximum of %d",
keyLength, FBUtilities.MAX_UNSIGNED_SHORT));
}
ByteBuffer key = ByteBufferUtil.read(input, keyLength);
int generation = input.readInt();
input.readBoolean(); // backwards compatibility for "promoted indexes" boolean
SSTableReader reader = null;
if (!skipEntry)
{
Pair qualifiedName = Pair.create(cfs.metadata.ksName, cfs.metadata.cfName);
Map generationToSSTableReader = cachedSSTableReaders.get(qualifiedName);
if (generationToSSTableReader == null)
{
generationToSSTableReader = new HashMap<>(cfs.getLiveSSTables().size());
for (SSTableReader ssTableReader : cfs.getSSTables(SSTableSet.CANONICAL))
{
generationToSSTableReader.put(ssTableReader.descriptor.generation, ssTableReader);
}
cachedSSTableReaders.putIfAbsent(qualifiedName, generationToSSTableReader);
}
reader = generationToSSTableReader.get(generation);
}
if (skipEntry || reader == null)
{
// The sstable doesn't exist anymore, so we can't be sure of the exact version and assume its the current version. The only case where we'll be
// wrong is during upgrade, in which case we fail at deserialization. This is not a huge deal however since 1) this is unlikely enough that
// this won't affect many users (if any) and only once, 2) this doesn't prevent the node from starting and 3) CASSANDRA-10219 shows that this
// part of the code has been broken for a while without anyone noticing (it is, btw, still broken until CASSANDRA-10219 is fixed).
RowIndexEntry.Serializer.skip(input, BigFormat.instance.getLatestVersion());
return null;
}
RowIndexEntry.IndexSerializer indexSerializer = reader.descriptor.getFormat().getIndexSerializer(reader.metadata,
reader.descriptor.version,
SerializationHeader.forKeyCache(cfs.metadata));
RowIndexEntry entry = indexSerializer.deserialize(input);
return Futures.immediateFuture(Pair.create(new KeyCacheKey(cfs.metadata.ksAndCFName, reader.descriptor, key), entry));
}
public void cleanupAfterDeserialize()
{
cachedSSTableReaders.clear();
}
}
}
© 2015 - 2024 Weber Informatics LLC | Privacy Policy