ByteBuffer createBuffer(CommitLog commitLog) {
try {
// Extend the file size to the standard segment size.
// NOTE: while we're using RAF to easily adjust file size, we need to avoid using RAF
// for grabbing the FileChannel due to FILE_SHARE_DELETE flag bug on windows.
// See: https://bugs.openjdk.java.net/browse/JDK-6357433 and CASSANDRA-8308
try (RandomAccessFile raf = new RandomAccessFile(logFile, "rw")) {
raf.setLength(DatabaseDescriptor.getCommitLogSegmentSize());
} catch (IOException e) {
throw new FSWriteError(e, logFile);
}
commitLog.allocator.addSize(DatabaseDescriptor.getCommitLogSegmentSize());
return channel.map(
FileChannel.MapMode.READ_WRITE, 0, DatabaseDescriptor.getCommitLogSegmentSize());
} catch (IOException e) {
throw new FSWriteError(e, logFile);
}
}
/**
* param @ table - name of table for which we are maintaining this commit log. param @
* recoverymode - is commit log being instantiated in in recovery mode.
*/
private CommitLog() {
try {
DatabaseDescriptor.createAllDirectories();
segmentSize = DatabaseDescriptor.getCommitLogSegmentSize();
} catch (IOException e) {
throw new IOError(e);
}
// all old segments are recovered and deleted before CommitLog is instantiated.
// All we need to do is create a new one.
segments.add(new CommitLogSegment());
executor =
DatabaseDescriptor.getCommitLogSync() == Config.CommitLogSync.batch
? new BatchCommitLogExecutorService()
: new PeriodicCommitLogExecutorService(this);
}
public void makeLog() throws IOException, InterruptedException {
CommitLog commitLog = CommitLog.instance;
System.out.format(
"\nUsing commit log size %dmb, compressor %s, sync %s%s\n",
mb(DatabaseDescriptor.getCommitLogSegmentSize()),
commitLog.configuration.getCompressorName(),
commitLog.executor.getClass().getSimpleName(),
randomSize ? " random size" : "");
final List<CommitlogExecutor> threads = new ArrayList<>();
ScheduledExecutorService scheduled = startThreads(commitLog, threads);
Thread.sleep(runTimeMs);
stop = true;
scheduled.shutdown();
scheduled.awaitTermination(2, TimeUnit.SECONDS);
int hash = 0;
int cells = 0;
for (CommitlogExecutor t : threads) {
t.join();
hash += t.hash;
cells += t.cells;
}
commitLog.shutdownBlocking();
File dataDir = new File(CommitLogUpgradeTest.DATA_DIR + FBUtilities.getReleaseVersionString());
System.out.format("Data will be stored in %s\n", dataDir);
if (dataDir.exists()) FileUtils.deleteRecursive(dataDir);
dataDir.mkdirs();
for (File f : new File(DatabaseDescriptor.getCommitLogLocation()).listFiles())
FileUtils.createHardLink(f, new File(dataDir, f.getName()));
Properties prop = new Properties();
prop.setProperty(CFID_PROPERTY, Schema.instance.getId(KEYSPACE, TABLE).toString());
prop.setProperty(CELLS_PROPERTY, Integer.toString(cells));
prop.setProperty(HASH_PROPERTY, Integer.toString(hash));
prop.store(
new FileOutputStream(new File(dataDir, PROPERTIES_FILE)),
"CommitLog upgrade test, version " + FBUtilities.getReleaseVersionString());
System.out.println("Done");
}
/*
* Commit Log tracks every write operation into the system. The aim of the commit log is to be able to
* successfully recover data that was not stored to disk via the Memtable.
*/
public class CommitLog implements CommitLogMBean {
private static final Logger logger = LoggerFactory.getLogger(CommitLog.class);
public static final CommitLog instance = CommitLog.construct();
// we only permit records HALF the size of a commit log, to ensure we don't spin allocating many
// mostly
// empty segments when writing large records
private final long MAX_MUTATION_SIZE = DatabaseDescriptor.getCommitLogSegmentSize() >> 1;
public final CommitLogSegmentManager allocator;
public final CommitLogArchiver archiver;
final CommitLogMetrics metrics;
final AbstractCommitLogService executor;
final ICompressor compressor;
public ParameterizedClass compressorClass;
public final String location;
private static CommitLog construct() {
CommitLog log =
new CommitLog(DatabaseDescriptor.getCommitLogLocation(), CommitLogArchiver.construct());
MBeanServer mbs = ManagementFactory.getPlatformMBeanServer();
try {
mbs.registerMBean(log, new ObjectName("org.apache.cassandra.db:type=Commitlog"));
} catch (Exception e) {
throw new RuntimeException(e);
}
return log.start();
}
@VisibleForTesting
CommitLog(String location, CommitLogArchiver archiver) {
compressorClass = DatabaseDescriptor.getCommitLogCompression();
this.location = location;
ICompressor compressor =
compressorClass != null ? CompressionParameters.createCompressor(compressorClass) : null;
DatabaseDescriptor.createAllDirectories();
this.compressor = compressor;
this.archiver = archiver;
metrics = new CommitLogMetrics();
executor =
DatabaseDescriptor.getCommitLogSync() == Config.CommitLogSync.batch
? new BatchCommitLogService(this)
: new PeriodicCommitLogService(this);
allocator = new CommitLogSegmentManager(this);
// register metrics
metrics.attach(executor, allocator);
}
CommitLog start() {
executor.start();
allocator.start();
return this;
}
/**
* Perform recovery on commit logs located in the directory specified by the config file.
*
* @return the number of mutations replayed
*/
public int recover() throws IOException {
// If createReserveSegments is already flipped, the CLSM is running and recovery has already
// taken place.
if (allocator.createReserveSegments) return 0;
// Allocator could be in the process of initial startup with 0 active and available segments. We
// need to wait for
// the allocation manager to finish allocation and add it to available segments so we don't get
// an invalid response
// on allocator.manages(...) below by grabbing a file off the filesystem before it's added to
// the CLQ.
allocator.allocatingFrom();
FilenameFilter unmanagedFilesFilter =
new FilenameFilter() {
public boolean accept(File dir, String name) {
// we used to try to avoid instantiating commitlog (thus creating an empty segment ready
// for writes)
// until after recover was finished. this turns out to be fragile; it is less
// error-prone to go
// ahead and allow writes before recover(), and just skip active segments when we do.
return CommitLogDescriptor.isValid(name) && !allocator.manages(name);
}
};
// submit all existing files in the commit log dir for archiving prior to recovery -
// CASSANDRA-6904
for (File file :
new File(DatabaseDescriptor.getCommitLogLocation()).listFiles(unmanagedFilesFilter)) {
archiver.maybeArchive(file.getPath(), file.getName());
archiver.maybeWaitForArchiving(file.getName());
}
assert archiver.archivePending.isEmpty()
: "Not all commit log archive tasks were completed before restore";
archiver.maybeRestoreArchive();
File[] files =
new File(DatabaseDescriptor.getCommitLogLocation()).listFiles(unmanagedFilesFilter);
int replayed = 0;
if (files.length == 0) {
logger.info("No commitlog files found; skipping replay");
} else {
Arrays.sort(files, new CommitLogSegmentFileComparator());
logger.info("Replaying {}", StringUtils.join(files, ", "));
replayed = recover(files);
logger.info("Log replay complete, {} replayed mutations", replayed);
for (File f : files) allocator.recycleSegment(f);
}
allocator.enableReserveSegmentCreation();
return replayed;
}
/**
* Perform recovery on a list of commit log files.
*
* @param clogs the list of commit log files to replay
* @return the number of mutations replayed
*/
public int recover(File... clogs) throws IOException {
CommitLogReplayer recovery = CommitLogReplayer.construct(this);
recovery.recover(clogs);
return recovery.blockForWrites();
}
/** Perform recovery on a single commit log. */
public void recover(String path) throws IOException {
CommitLogReplayer recovery = CommitLogReplayer.construct(this);
recovery.recover(new File(path), false);
recovery.blockForWrites();
}
/**
* @return a ReplayPosition which, if >= one returned from add(), implies add() was started (but
* not necessarily finished) prior to this call
*/
public ReplayPosition getContext() {
return allocator.allocatingFrom().getContext();
}
/**
* Flushes all dirty CFs, waiting for them to free and recycle any segments they were retaining
*/
public void forceRecycleAllSegments(Iterable<UUID> droppedCfs) {
allocator.forceRecycleAll(droppedCfs);
}
/**
* Flushes all dirty CFs, waiting for them to free and recycle any segments they were retaining
*/
public void forceRecycleAllSegments() {
allocator.forceRecycleAll(Collections.<UUID>emptyList());
}
/** Forces a disk flush on the commit log files that need it. Blocking. */
public void sync(boolean syncAllSegments) {
CommitLogSegment current = allocator.allocatingFrom();
for (CommitLogSegment segment : allocator.getActiveSegments()) {
if (!syncAllSegments && segment.id > current.id) return;
segment.sync();
}
}
/** Preempts the CLExecutor, telling to to sync immediately */
public void requestExtraSync() {
executor.requestExtraSync();
}
/**
* Add a Mutation to the commit log.
*
* @param mutation the Mutation to add to the log
*/
public ReplayPosition add(Mutation mutation) {
assert mutation != null;
long size = Mutation.serializer.serializedSize(mutation, MessagingService.current_version);
long totalSize = size + ENTRY_OVERHEAD_SIZE;
if (totalSize > MAX_MUTATION_SIZE) {
throw new IllegalArgumentException(
String.format(
"Mutation of %s bytes is too large for the maxiumum size of %s",
totalSize, MAX_MUTATION_SIZE));
}
Allocation alloc = allocator.allocate(mutation, (int) totalSize);
ICRC32 checksum = CRC32Factory.instance.create();
final ByteBuffer buffer = alloc.getBuffer();
try (BufferedDataOutputStreamPlus dos = new DataOutputBufferFixed(buffer)) {
// checksummed length
dos.writeInt((int) size);
checksum.update(buffer, buffer.position() - 4, 4);
buffer.putInt(checksum.getCrc());
int start = buffer.position();
// checksummed mutation
Mutation.serializer.serialize(mutation, dos, MessagingService.current_version);
checksum.update(buffer, start, (int) size);
buffer.putInt(checksum.getCrc());
} catch (IOException e) {
throw new FSWriteError(e, alloc.getSegment().getPath());
} finally {
alloc.markWritten();
}
executor.finishWriteFor(alloc);
return alloc.getReplayPosition();
}
/**
* Modifies the per-CF dirty cursors of any commit log segments for the column family according to
* the position given. Discards any commit log segments that are no longer used.
*
* @param cfId the column family ID that was flushed
* @param context the replay position of the flush
*/
public void discardCompletedSegments(final UUID cfId, final ReplayPosition context) {
logger.trace("discard completed log segments for {}, table {}", context, cfId);
// Go thru the active segment files, which are ordered oldest to newest, marking the
// flushed CF as clean, until we reach the segment file containing the ReplayPosition passed
// in the arguments. Any segments that become unused after they are marked clean will be
// recycled or discarded.
for (Iterator<CommitLogSegment> iter = allocator.getActiveSegments().iterator();
iter.hasNext(); ) {
CommitLogSegment segment = iter.next();
segment.markClean(cfId, context);
if (segment.isUnused()) {
logger.trace("Commit log segment {} is unused", segment);
allocator.recycleSegment(segment);
} else {
logger.trace(
"Not safe to delete{} commit log segment {}; dirty is {}",
(iter.hasNext() ? "" : " active"),
segment,
segment.dirtyString());
}
// Don't mark or try to delete any newer segments once we've reached the one containing the
// position of the flush.
if (segment.contains(context)) break;
}
}
@Override
public String getArchiveCommand() {
return archiver.archiveCommand;
}
@Override
public String getRestoreCommand() {
return archiver.restoreCommand;
}
@Override
public String getRestoreDirectories() {
return archiver.restoreDirectories;
}
@Override
public long getRestorePointInTime() {
return archiver.restorePointInTime;
}
@Override
public String getRestorePrecision() {
return archiver.precision.toString();
}
public List<String> getActiveSegmentNames() {
List<String> segmentNames = new ArrayList<>();
for (CommitLogSegment segment : allocator.getActiveSegments())
segmentNames.add(segment.getName());
return segmentNames;
}
public List<String> getArchivingSegmentNames() {
return new ArrayList<>(archiver.archivePending.keySet());
}
@Override
public long getActiveContentSize() {
long size = 0;
for (CommitLogSegment segment : allocator.getActiveSegments()) size += segment.contentSize();
return size;
}
@Override
public long getActiveOnDiskSize() {
return allocator.onDiskSize();
}
@Override
public Map<String, Double> getActiveSegmentCompressionRatios() {
Map<String, Double> segmentRatios = new TreeMap<>();
for (CommitLogSegment segment : allocator.getActiveSegments())
segmentRatios.put(segment.getName(), 1.0 * segment.onDiskSize() / segment.contentSize());
return segmentRatios;
}
/** Shuts down the threads used by the commit log, blocking until completion. */
public void shutdownBlocking() throws InterruptedException {
executor.shutdown();
executor.awaitTermination();
allocator.shutdown();
allocator.awaitTermination();
}
/**
* FOR TESTING PURPOSES. See CommitLogAllocator.
*
* @return the number of files recovered
*/
public int resetUnsafe(boolean deleteSegments) throws IOException {
stopUnsafe(deleteSegments);
return restartUnsafe();
}
/** FOR TESTING PURPOSES. See CommitLogAllocator. */
public void stopUnsafe(boolean deleteSegments) {
executor.shutdown();
try {
executor.awaitTermination();
} catch (InterruptedException e) {
throw new RuntimeException(e);
}
allocator.stopUnsafe(deleteSegments);
}
/** FOR TESTING PURPOSES. See CommitLogAllocator */
public int restartUnsafe() throws IOException {
allocator.start();
executor.restartUnsafe();
try {
return recover();
} catch (FSWriteError e) {
// Workaround for a class of races that keeps showing up on Windows tests.
// stop/start/reset path on Windows with segment deletion is very touchy/brittle
// and the timing keeps getting screwed up. Rather than chasing our tail further
// or rewriting the CLSM, just report that we didn't recover anything back up
// the chain. This will silence most intermittent test failures on Windows
// and appropriately fail tests that expected segments to be recovered that
// were not.
return 0;
}
}
/**
* Used by tests.
*
* @return the number of active segments (segments with unflushed data in them)
*/
public int activeSegments() {
return allocator.getActiveSegments().size();
}
@VisibleForTesting
public static boolean handleCommitError(String message, Throwable t) {
JVMStabilityInspector.inspectCommitLogThrowable(t);
switch (DatabaseDescriptor.getCommitFailurePolicy()) {
// Needed here for unit tests to not fail on default assertion
case die:
case stop:
StorageService.instance.stopTransports();
// $FALL-THROUGH$
case stop_commit:
logger.error(
String.format(
"%s. Commit disk failure policy is %s; terminating thread",
message, DatabaseDescriptor.getCommitFailurePolicy()),
t);
return false;
case ignore:
logger.error(message, t);
return true;
default:
throw new AssertionError(DatabaseDescriptor.getCommitFailurePolicy());
}
}
}
@Override
public long onDiskSize() {
return DatabaseDescriptor.getCommitLogSegmentSize();
}