/*
* 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.db.compaction;
import java.io.File;
import java.io.IOException;
import java.util.*;
import java.util.concurrent.TimeUnit;
import com.google.common.base.*;
import com.google.common.collect.*;
import org.apache.commons.lang3.StringUtils;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.apache.cassandra.config.DatabaseDescriptor;
import org.apache.cassandra.db.ColumnFamilyStore;
import org.apache.cassandra.db.DecoratedKey;
import org.apache.cassandra.db.RowIndexEntry;
import org.apache.cassandra.db.SystemKeyspace;
import org.apache.cassandra.db.compaction.CompactionManager.CompactionExecutorStatsCollector;
import org.apache.cassandra.io.sstable.Descriptor;
import org.apache.cassandra.io.sstable.SSTableReader;
import org.apache.cassandra.io.sstable.SSTableWriter;
import org.apache.cassandra.io.sstable.metadata.MetadataCollector;
import org.apache.cassandra.service.ActiveRepairService;
import org.apache.cassandra.utils.CloseableIterator;
public class CompactionTask extends AbstractCompactionTask
{
protected static final Logger logger = LoggerFactory.getLogger(CompactionTask.class);
protected final int gcBefore;
protected static long totalBytesCompacted = 0;
private Set<SSTableReader> toCompact;
private CompactionExecutorStatsCollector collector;
public CompactionTask(ColumnFamilyStore cfs, Iterable<SSTableReader> sstables, final int gcBefore)
{
super(cfs, sstables);
this.gcBefore = gcBefore;
toCompact = Sets.newHashSet(sstables);
}
public static synchronized long addToTotalBytesCompacted(long bytesCompacted)
{
return totalBytesCompacted += bytesCompacted;
}
protected int executeInternal(CompactionExecutorStatsCollector collector)
{
this.collector = collector;
run();
return toCompact.size();
}
public long getExpectedWriteSize()
{
return cfs.getExpectedCompactedFileSize(toCompact, compactionType);
}
public boolean reduceScopeForLimitedSpace()
{
if (partialCompactionsAcceptable() && toCompact.size() > 1)
{
// Try again w/o the largest one.
logger.warn("insufficient space to compact all requested files {}", StringUtils.join(toCompact, ", "));
// Note that we have removed files that are still marked as compacting.
// This suboptimal but ok since the caller will unmark all the sstables at the end.
return toCompact.remove(cfs.getMaxSizeFile(toCompact));
}
else
{
return false;
}
}
/**
* For internal use and testing only. The rest of the system should go through the submit* methods,
* which are properly serialized.
* Caller is in charge of marking/unmarking the sstables as compacting.
*/
protected void runWith(File sstableDirectory) throws Exception
{
// The collection of sstables passed may be empty (but not null); even if
// it is not empty, it may compact down to nothing if all rows are deleted.
assert sstables != null && sstableDirectory != null;
if (toCompact.size() == 0)
return;
// Note that the current compaction strategy, is not necessarily the one this task was created under.
// This should be harmless; see comments to CFS.maybeReloadCompactionStrategy.
AbstractCompactionStrategy strategy = cfs.getCompactionStrategy();
if (DatabaseDescriptor.isSnapshotBeforeCompaction())
cfs.snapshotWithoutFlush(System.currentTimeMillis() + "-compact-" + cfs.name);
// sanity check: all sstables must belong to the same cfs
assert !Iterables.any(toCompact, new Predicate<SSTableReader>()
{
@Override
public boolean apply(SSTableReader sstable)
{
return !sstable.descriptor.cfname.equals(cfs.name);
}
});
UUID taskId = SystemKeyspace.startCompaction(cfs, toCompact);
CompactionController controller = getCompactionController(toCompact);
Set<SSTableReader> actuallyCompact = Sets.difference(toCompact, controller.getFullyExpiredSSTables());
// new sstables from flush can be added during a compaction, but only the compaction can remove them,
// so in our single-threaded compaction world this is a valid way of determining if we're compacting
// all the sstables (that existed when we started)
logger.info("Compacting {}", toCompact);
long start = System.nanoTime();
long totalKeysWritten = 0;
long estimatedTotalKeys = Math.max(cfs.metadata.getMinIndexInterval(), SSTableReader.getApproximateKeyCount(actuallyCompact));
long estimatedSSTables = Math.max(1, SSTableReader.getTotalBytes(actuallyCompact) / strategy.getMaxSSTableBytes());
long keysPerSSTable = (long) Math.ceil((double) estimatedTotalKeys / estimatedSSTables);
logger.debug("Expected bloom filter size : {}", keysPerSSTable);
AbstractCompactionIterable ci = new CompactionIterable(compactionType, strategy.getScanners(actuallyCompact), controller);
CloseableIterator<AbstractCompactedRow> iter = ci.iterator();
Map<DecoratedKey, RowIndexEntry> cachedKeys = new HashMap<>();
// we can't preheat until the tracker has been set. This doesn't happen until we tell the cfs to
// replace the old entries. Track entries to preheat here until then.
Map<Descriptor, Map<DecoratedKey, RowIndexEntry>> cachedKeyMap = new HashMap<>();
Collection<SSTableReader> sstables = new ArrayList<>();
Collection<SSTableWriter> writers = new ArrayList<>();
long minRepairedAt = getMinRepairedAt(actuallyCompact);
if (collector != null)
collector.beginCompaction(ci);
try
{
if (!iter.hasNext())
{
// don't mark compacted in the finally block, since if there _is_ nondeleted data,
// we need to sync it (via closeAndOpen) first, so there is no period during which
// a crash could cause data loss.
cfs.markObsolete(toCompact, compactionType);
return;
}
SSTableWriter writer = createCompactionWriter(sstableDirectory, keysPerSSTable, minRepairedAt);
writers.add(writer);
while (iter.hasNext())
{
if (ci.isStopRequested())
throw new CompactionInterruptedException(ci.getCompactionInfo());
AbstractCompactedRow row = iter.next();
RowIndexEntry indexEntry = writer.append(row);
if (indexEntry == null)
{
controller.invalidateCachedRow(row.key);
row.close();
continue;
}
totalKeysWritten++;
if (DatabaseDescriptor.getPreheatKeyCache())
{
for (SSTableReader sstable : actuallyCompact)
{
if (sstable.getCachedPosition(row.key, false) != null)
{
cachedKeys.put(row.key, indexEntry);
break;
}
}
}
if (newSSTableSegmentThresholdReached(writer))
{
// tmp = false because later we want to query it with descriptor from SSTableReader
cachedKeyMap.put(writer.descriptor.asTemporary(false), cachedKeys);
writer = createCompactionWriter(sstableDirectory, keysPerSSTable, minRepairedAt);
writers.add(writer);
cachedKeys = new HashMap<>();
}
}
if (writer.getFilePointer() > 0)
{
cachedKeyMap.put(writer.descriptor.asTemporary(false), cachedKeys);
}
else
{
writer.abort();
writers.remove(writer);
}
long maxAge = getMaxDataAge(toCompact);
for (SSTableWriter completedWriter : writers)
sstables.add(completedWriter.closeAndOpenReader(maxAge));
}
catch (Throwable t)
{
for (SSTableWriter writer : writers)
writer.abort();
// also remove already completed SSTables
for (SSTableReader sstable : sstables)
{
sstable.markObsolete();
sstable.releaseReference();
}
throw Throwables.propagate(t);
}
finally
{
controller.close();
// point of no return -- the new sstables are live on disk; next we'll start deleting the old ones
// (in replaceCompactedSSTables)
if (taskId != null)
SystemKeyspace.finishCompaction(taskId);
if (collector != null)
collector.finishCompaction(ci);
try
{
// We don't expect this to throw, but just in case, we do it after the cleanup above, to make sure
// we don't end up with compaction information hanging around indefinitely in limbo.
iter.close();
}
catch (IOException e)
{
throw new RuntimeException(e);
}
}
replaceCompactedSSTables(toCompact, sstables);
// TODO: this doesn't belong here, it should be part of the reader to load when the tracker is wired up
for (SSTableReader sstable : sstables)
sstable.preheat(cachedKeyMap.get(sstable.descriptor));
// log a bunch of statistics about the result and save to system table compaction_history
long dTime = TimeUnit.NANOSECONDS.toMillis(System.nanoTime() - start);
long startsize = SSTableReader.getTotalBytes(toCompact);
long endsize = SSTableReader.getTotalBytes(sstables);
double ratio = (double) endsize / (double) startsize;
StringBuilder builder = new StringBuilder();
for (SSTableReader reader : sstables)
builder.append(reader.descriptor.baseFilename()).append(",");
double mbps = dTime > 0 ? (double) endsize / (1024 * 1024) / ((double) dTime / 1000) : 0;
long totalSourceRows = 0;
long[] counts = ci.getMergedRowCounts();
StringBuilder mergeSummary = new StringBuilder(counts.length * 10);
Map<Integer, Long> mergedRows = new HashMap<>();
for (int i = 0; i < counts.length; i++)
{
long count = counts[i];
if (count == 0)
continue;
int rows = i + 1;
totalSourceRows += rows * count;
mergeSummary.append(String.format("%d:%d, ", rows, count));
mergedRows.put(rows, count);
}
SystemKeyspace.updateCompactionHistory(cfs.keyspace.getName(), cfs.name, start, startsize, endsize, mergedRows);
logger.info(String.format("Compacted %d sstables to [%s]. %,d bytes to %,d (~%d%% of original) in %,dms = %fMB/s. %,d total partitions merged to %,d. Partition merge counts were {%s}",
toCompact.size(), builder.toString(), startsize, endsize, (int) (ratio * 100), dTime, mbps, totalSourceRows, totalKeysWritten, mergeSummary.toString()));
logger.debug(String.format("CF Total Bytes Compacted: %,d", CompactionTask.addToTotalBytesCompacted(endsize)));
logger.debug("Actual #keys: {}, Estimated #keys:{}, Err%: {}", totalKeysWritten, estimatedTotalKeys, ((double)(totalKeysWritten - estimatedTotalKeys)/totalKeysWritten));
}
private long getMinRepairedAt(Set<SSTableReader> actuallyCompact)
{
long minRepairedAt= Long.MAX_VALUE;
for (SSTableReader sstable : actuallyCompact)
minRepairedAt = Math.min(minRepairedAt, sstable.getSSTableMetadata().repairedAt);
if (minRepairedAt == Long.MAX_VALUE)
return ActiveRepairService.UNREPAIRED_SSTABLE;
return minRepairedAt;
}
private SSTableWriter createCompactionWriter(File sstableDirectory, long keysPerSSTable, long repairedAt)
{
return new SSTableWriter(cfs.getTempSSTablePath(sstableDirectory),
keysPerSSTable,
repairedAt,
cfs.metadata,
cfs.partitioner,
new MetadataCollector(toCompact, cfs.metadata.comparator, getLevel()));
}
protected int getLevel()
{
return 0;
}
protected void replaceCompactedSSTables(Collection<SSTableReader> compacted, Collection<SSTableReader> replacements)
{
cfs.replaceCompactedSSTables(compacted, replacements, compactionType);
}
protected CompactionController getCompactionController(Set<SSTableReader> toCompact)
{
return new CompactionController(cfs, toCompact, gcBefore);
}
protected boolean partialCompactionsAcceptable()
{
return !isUserDefined;
}
// extensibility point for other strategies that may want to limit the upper bounds of the sstable segment size
protected boolean newSSTableSegmentThresholdReached(SSTableWriter writer)
{
return false;
}
public static long getMaxDataAge(Collection<SSTableReader> sstables)
{
long max = 0;
for (SSTableReader sstable : sstables)
{
if (sstable.maxDataAge > max)
max = sstable.maxDataAge;
}
return max;
}
}