LevelDB 1.13

Fix issues 77, 87, 182, 190.

Additionally, fix the bug described in
https://groups.google.com/d/msg/leveldb/yL6h1mAOc20/vLU64RylIdMJ
where a large contiguous keyspace of deleted data was not getting
compacted.

Also fix a bug where options.max_open_files was not getting clamped
properly.

Makefile

@@ -31,6 +31,7 @@ TESTHARNESS = ./util/testharness.o $(TESTUTIL)
 
 TESTS = \
 	arena_test \
+	autocompact_test \
 	bloom_test \
 	c_test \
 	cache_test \
@@ -70,7 +71,7 @@ SHARED = $(SHARED1)
 else
 # Update db.h if you change these.
 SHARED_MAJOR = 1
-SHARED_MINOR = 12
+SHARED_MINOR = 13
 SHARED1 = libleveldb.$(PLATFORM_SHARED_EXT)
 SHARED2 = $(SHARED1).$(SHARED_MAJOR)
 SHARED3 = $(SHARED1).$(SHARED_MAJOR).$(SHARED_MINOR)
@@ -114,6 +115,9 @@ leveldbutil: db/leveldb_main.o $(LIBOBJECTS)
 arena_test: util/arena_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(CXX) $(LDFLAGS) util/arena_test.o $(LIBOBJECTS) $(TESTHARNESS) -o $@ $(LIBS)
 
+autocompact_test: db/autocompact_test.o $(LIBOBJECTS) $(TESTHARNESS)
+	$(CXX) $(LDFLAGS) db/autocompact_test.o $(LIBOBJECTS) $(TESTHARNESS) -o $@ $(LIBS)
+
 bloom_test: util/bloom_test.o $(LIBOBJECTS) $(TESTHARNESS)
 	$(CXX) $(LDFLAGS) util/bloom_test.o $(LIBOBJECTS) $(TESTHARNESS) -o $@ $(LIBS)
 

db/autocompact_test.cc

@@ -0,0 +1,118 @@
+// Copyright (c) 2013 The LevelDB Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style license that can be
+// found in the LICENSE file. See the AUTHORS file for names of contributors.
+
+#include "leveldb/db.h"
+#include "db/db_impl.h"
+#include "leveldb/cache.h"
+#include "util/testharness.h"
+#include "util/testutil.h"
+
+namespace leveldb {
+
+class AutoCompactTest {
+ public:
+  std::string dbname_;
+  Cache* tiny_cache_;
+  Options options_;
+  DB* db_;
+
+  AutoCompactTest() {
+    dbname_ = test::TmpDir() + "/autocompact_test";
+    tiny_cache_ = NewLRUCache(100);
+    options_.block_cache = tiny_cache_;
+    DestroyDB(dbname_, options_);
+    options_.create_if_missing = true;
+    options_.compression = kNoCompression;
+    ASSERT_OK(DB::Open(options_, dbname_, &db_));
+  }
+
+  ~AutoCompactTest() {
+    delete db_;
+    DestroyDB(dbname_, Options());
+    delete tiny_cache_;
+  }
+
+  std::string Key(int i) {
+    char buf[100];
+    snprintf(buf, sizeof(buf), "key%06d", i);
+    return std::string(buf);
+  }
+
+  uint64_t Size(const Slice& start, const Slice& limit) {
+    Range r(start, limit);
+    uint64_t size;
+    db_->GetApproximateSizes(&r, 1, &size);
+    return size;
+  }
+
+  void DoReads(int n);
+};
+
+static const int kValueSize = 200 * 1024;
+static const int kTotalSize = 100 * 1024 * 1024;
+static const int kCount = kTotalSize / kValueSize;
+
+// Read through the first n keys repeatedly and check that they get
+// compacted (verified by checking the size of the key space).
+void AutoCompactTest::DoReads(int n) {
+  std::string value(kValueSize, 'x');
+  DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
+
+  // Fill database
+  for (int i = 0; i < kCount; i++) {
+    ASSERT_OK(db_->Put(WriteOptions(), Key(i), value));
+  }
+  ASSERT_OK(dbi->TEST_CompactMemTable());
+
+  // Delete everything
+  for (int i = 0; i < kCount; i++) {
+    ASSERT_OK(db_->Delete(WriteOptions(), Key(i)));
+  }
+  ASSERT_OK(dbi->TEST_CompactMemTable());
+
+  // Get initial measurement of the space we will be reading.
+  const int64_t initial_size = Size(Key(0), Key(n));
+  const int64_t initial_other_size = Size(Key(n), Key(kCount));
+
+  // Read until size drops significantly.
+  std::string limit_key = Key(n);
+  for (int read = 0; true; read++) {
+    ASSERT_LT(read, 100) << "Taking too long to compact";
+    Iterator* iter = db_->NewIterator(ReadOptions());
+    for (iter->SeekToFirst();
+         iter->Valid() && iter->key().ToString() < limit_key;
+         iter->Next()) {
+      // Drop data
+    }
+    delete iter;
+    // Wait a little bit to allow any triggered compactions to complete.
+    Env::Default()->SleepForMicroseconds(1000000);
+    uint64_t size = Size(Key(0), Key(n));
+    fprintf(stderr, "iter %3d => %7.3f MB [other %7.3f MB]\n",
+            read+1, size/1048576.0, Size(Key(n), Key(kCount))/1048576.0);
+    if (size <= initial_size/10) {
+      break;
+    }
+  }
+
+  // Verify that the size of the key space not touched by the reads
+  // is pretty much unchanged.
+  const int64_t final_other_size = Size(Key(n), Key(kCount));
+  ASSERT_LE(final_other_size, initial_other_size + 1048576);
+  ASSERT_GE(final_other_size, initial_other_size/5 - 1048576);
+}
+
+TEST(AutoCompactTest, ReadAll) {
+  DoReads(kCount);
+}
+
+TEST(AutoCompactTest, ReadHalf) {
+  DoReads(kCount/2);
+}
+
+}  // namespace leveldb
+
+int main(int argc, char** argv) {
+  return leveldb::test::RunAllTests();
+}

db/corruption_test.cc

@@ -35,6 +35,7 @@ class CorruptionTest {
   CorruptionTest() {
     tiny_cache_ = NewLRUCache(100);
     options_.env = &env_;
+    options_.block_cache = tiny_cache_;
     dbname_ = test::TmpDir() + "/db_test";
     DestroyDB(dbname_, options_);
 
@@ -50,17 +51,14 @@ class CorruptionTest {
      delete tiny_cache_;
   }
 
-  Status TryReopen(Options* options = NULL) {
+  Status TryReopen() {
     delete db_;
     db_ = NULL;
-    Options opt = (options ? *options : options_);
-    opt.env = &env_;
-    opt.block_cache = tiny_cache_;
-    return DB::Open(opt, dbname_, &db_);
+    return DB::Open(options_, dbname_, &db_);
   }
 
-  void Reopen(Options* options = NULL) {
-    ASSERT_OK(TryReopen(options));
+  void Reopen() {
+    ASSERT_OK(TryReopen());
   }
 
   void RepairDB() {
@@ -92,6 +90,10 @@ class CorruptionTest {
     for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
       uint64_t key;
       Slice in(iter->key());
+      if (in == "" || in == "~") {
+        // Ignore boundary keys.
+        continue;
+      }
       if (!ConsumeDecimalNumber(&in, &key) ||
           !in.empty() ||
           key < next_expected) {
@@ -233,7 +235,7 @@ TEST(CorruptionTest, TableFile) {
   dbi->TEST_CompactRange(1, NULL, NULL);
 
   Corrupt(kTableFile, 100, 1);
-  Check(99, 99);
+  Check(90, 99);
 }
 
 TEST(CorruptionTest, TableFileIndexData) {
@@ -299,7 +301,7 @@ TEST(CorruptionTest, CompactionInputError) {
   ASSERT_EQ(1, Property("leveldb.num-files-at-level" + NumberToString(last)));
 
   Corrupt(kTableFile, 100, 1);
-  Check(9, 9);
+  Check(5, 9);
 
   // Force compactions by writing lots of values
   Build(10000);
@@ -307,32 +309,23 @@ TEST(CorruptionTest, CompactionInputError) {
 }
 
 TEST(CorruptionTest, CompactionInputErrorParanoid) {
-  Options options;
-  options.paranoid_checks = true;
-  options.write_buffer_size = 1048576;
-  Reopen(&options);
+  options_.paranoid_checks = true;
+  options_.write_buffer_size = 512 << 10;
+  Reopen();
   DBImpl* dbi = reinterpret_cast<DBImpl*>(db_);
 
-  // Fill levels >= 1 so memtable compaction outputs to level 1
-  for (int level = 1; level < config::kNumLevels; level++) {
-    dbi->Put(WriteOptions(), "", "begin");
-    dbi->Put(WriteOptions(), "~", "end");
+  // Make multiple inputs so we need to compact.
+  for (int i = 0; i < 2; i++) {
+    Build(10);
     dbi->TEST_CompactMemTable();
+    Corrupt(kTableFile, 100, 1);
+    env_.SleepForMicroseconds(100000);
   }
+  dbi->CompactRange(NULL, NULL);
 
-  Build(10);
-  dbi->TEST_CompactMemTable();
-  ASSERT_EQ(1, Property("leveldb.num-files-at-level0"));
-
-  Corrupt(kTableFile, 100, 1);
-  Check(9, 9);
-
-  // Write must eventually fail because of corrupted table
-  Status s;
+  // Write must fail because of corrupted table
   std::string tmp1, tmp2;
-  for (int i = 0; i < 10000 && s.ok(); i++) {
-    s = db_->Put(WriteOptions(), Key(i, &tmp1), Value(i, &tmp2));
-  }
+  Status s = db_->Put(WriteOptions(), Key(5, &tmp1), Value(5, &tmp2));
   ASSERT_TRUE(!s.ok()) << "write did not fail in corrupted paranoid db";
 }
 

db/db_impl.cc

@@ -113,14 +113,14 @@ Options SanitizeOptions(const std::string& dbname,
   return result;
 }
 
-DBImpl::DBImpl(const Options& options, const std::string& dbname)
-    : env_(options.env),
-      internal_comparator_(options.comparator),
-      internal_filter_policy_(options.filter_policy),
-      options_(SanitizeOptions(
-          dbname, &internal_comparator_, &internal_filter_policy_, options)),
-      owns_info_log_(options_.info_log != options.info_log),
-      owns_cache_(options_.block_cache != options.block_cache),
+DBImpl::DBImpl(const Options& raw_options, const std::string& dbname)
+    : env_(raw_options.env),
+      internal_comparator_(raw_options.comparator),
+      internal_filter_policy_(raw_options.filter_policy),
+      options_(SanitizeOptions(dbname, &internal_comparator_,
+                               &internal_filter_policy_, raw_options)),
+      owns_info_log_(options_.info_log != raw_options.info_log),
+      owns_cache_(options_.block_cache != raw_options.block_cache),
       dbname_(dbname),
       db_lock_(NULL),
       shutting_down_(NULL),
@@ -130,6 +130,7 @@ DBImpl::DBImpl(const Options& options, const std::string& dbname)
       logfile_(NULL),
       logfile_number_(0),
       log_(NULL),
+      seed_(0),
       tmp_batch_(new WriteBatch),
       bg_compaction_scheduled_(false),
       manual_compaction_(NULL),
@@ -138,7 +139,7 @@ DBImpl::DBImpl(const Options& options, const std::string& dbname)
   has_imm_.Release_Store(NULL);
 
   // Reserve ten files or so for other uses and give the rest to TableCache.
-  const int table_cache_size = options.max_open_files - kNumNonTableCacheFiles;
+  const int table_cache_size = options_.max_open_files - kNumNonTableCacheFiles;
   table_cache_ = new TableCache(dbname_, &options_, table_cache_size);
 
   versions_ = new VersionSet(dbname_, &options_, table_cache_,
@@ -1027,7 +1028,8 @@ static void CleanupIteratorState(void* arg1, void* arg2) {
 }  // namespace
 
 Iterator* DBImpl::NewInternalIterator(const ReadOptions& options,
-                                      SequenceNumber* latest_snapshot) {
+                                      SequenceNumber* latest_snapshot,
+                                      uint32_t* seed) {
   IterState* cleanup = new IterState;
   mutex_.Lock();
   *latest_snapshot = versions_->LastSequence();
@@ -1051,13 +1053,15 @@ Iterator* DBImpl::NewInternalIterator(const ReadOptions& options,
   cleanup->version = versions_->current();
   internal_iter->RegisterCleanup(CleanupIteratorState, cleanup, NULL);
 
+  *seed = ++seed_;
   mutex_.Unlock();
   return internal_iter;
 }
 
 Iterator* DBImpl::TEST_NewInternalIterator() {
   SequenceNumber ignored;
-  return NewInternalIterator(ReadOptions(), &ignored);
+  uint32_t ignored_seed;
+  return NewInternalIterator(ReadOptions(), &ignored, &ignored_seed);
 }
 
 int64_t DBImpl::TEST_MaxNextLevelOverlappingBytes() {
@@ -1114,12 +1118,21 @@ Status DBImpl::Get(const ReadOptions& options,
 
 Iterator* DBImpl::NewIterator(const ReadOptions& options) {
   SequenceNumber latest_snapshot;
-  Iterator* internal_iter = NewInternalIterator(options, &latest_snapshot);
+  uint32_t seed;
+  Iterator* iter = NewInternalIterator(options, &latest_snapshot, &seed);
   return NewDBIterator(
-      &dbname_, env_, user_comparator(), internal_iter,
+      this, user_comparator(), iter,
       (options.snapshot != NULL
        ? reinterpret_cast<const SnapshotImpl*>(options.snapshot)->number_
-       : latest_snapshot));
+       : latest_snapshot),
+      seed);
+}
+
+void DBImpl::RecordReadSample(Slice key) {
+  MutexLock l(&mutex_);
+  if (versions_->current()->RecordReadSample(key)) {
+    MaybeScheduleCompaction();
+  }
 }
 
 const Snapshot* DBImpl::GetSnapshot() {

db/db_impl.h

@@ -59,13 +59,19 @@ class DBImpl : public DB {
   // file at a level >= 1.
   int64_t TEST_MaxNextLevelOverlappingBytes();
 
+  // Record a sample of bytes read at the specified internal key.
+  // Samples are taken approximately once every config::kReadBytesPeriod
+  // bytes.
+  void RecordReadSample(Slice key);
+
  private:
   friend class DB;
   struct CompactionState;
   struct Writer;
 
   Iterator* NewInternalIterator(const ReadOptions&,
-                                SequenceNumber* latest_snapshot);
+                                SequenceNumber* latest_snapshot,
+                                uint32_t* seed);
 
   Status NewDB();
 
@@ -135,6 +141,7 @@ class DBImpl : public DB {
   WritableFile* logfile_;
   uint64_t logfile_number_;
   log::Writer* log_;
+  uint32_t seed_;                // For sampling.
 
   // Queue of writers.
   std::deque<Writer*> writers_;

db/db_iter.cc

@@ -5,12 +5,14 @@
 #include "db/db_iter.h"
 
 #include "db/filename.h"
+#include "db/db_impl.h"
 #include "db/dbformat.h"
 #include "leveldb/env.h"
 #include "leveldb/iterator.h"
 #include "port/port.h"
 #include "util/logging.h"
 #include "util/mutexlock.h"
+#include "util/random.h"
 
 namespace leveldb {
 
@@ -46,15 +48,16 @@ class DBIter: public Iterator {
     kReverse
   };
 
-  DBIter(const std::string* dbname, Env* env,
-         const Comparator* cmp, Iterator* iter, SequenceNumber s)
-      : dbname_(dbname),
-        env_(env),
+  DBIter(DBImpl* db, const Comparator* cmp, Iterator* iter, SequenceNumber s,
+         uint32_t seed)
+      : db_(db),
         user_comparator_(cmp),
         iter_(iter),
         sequence_(s),
         direction_(kForward),
-        valid_(false) {
+        valid_(false),
+        rnd_(seed),
+        bytes_counter_(RandomPeriod()) {
   }
   virtual ~DBIter() {
     delete iter_;
@@ -100,8 +103,12 @@ class DBIter: public Iterator {
     }
   }
 
-  const std::string* const dbname_;
-  Env* const env_;
+  // Pick next gap with average value of config::kReadBytesPeriod.
+  ssize_t RandomPeriod() {
+    return rnd_.Uniform(2*config::kReadBytesPeriod);
+  }
+
+  DBImpl* db_;
   const Comparator* const user_comparator_;
   Iterator* const iter_;
   SequenceNumber const sequence_;
@@ -112,13 +119,23 @@ class DBIter: public Iterator {
   Direction direction_;
   bool valid_;
 
+  Random rnd_;
+  ssize_t bytes_counter_;
+
   // No copying allowed
   DBIter(const DBIter&);
   void operator=(const DBIter&);
 };
 
 inline bool DBIter::ParseKey(ParsedInternalKey* ikey) {
-  if (!ParseInternalKey(iter_->key(), ikey)) {
+  Slice k = iter_->key();
+  ssize_t n = k.size() + iter_->value().size();
+  bytes_counter_ -= n;
+  while (bytes_counter_ < 0) {
+    bytes_counter_ += RandomPeriod();
+    db_->RecordReadSample(k);
+  }
+  if (!ParseInternalKey(k, ikey)) {
     status_ = Status::Corruption("corrupted internal key in DBIter");
     return false;
   } else {
@@ -288,12 +305,12 @@ void DBIter::SeekToLast() {
 }  // anonymous namespace
 
 Iterator* NewDBIterator(
-    const std::string* dbname,
-    Env* env,
+    DBImpl* db,
     const Comparator* user_key_comparator,
     Iterator* internal_iter,
-    const SequenceNumber& sequence) {
-  return new DBIter(dbname, env, user_key_comparator, internal_iter, sequence);
+    SequenceNumber sequence,
+    uint32_t seed) {
+  return new DBIter(db, user_key_comparator, internal_iter, sequence, seed);
 }
 
 }  // namespace leveldb

db/db_iter.h

@@ -11,15 +11,17 @@
 
 namespace leveldb {
 
+class DBImpl;
+
 // Return a new iterator that converts internal keys (yielded by
 // "*internal_iter") that were live at the specified "sequence" number
 // into appropriate user keys.
 extern Iterator* NewDBIterator(
-    const std::string* dbname,
-    Env* env,
+    DBImpl* db,
     const Comparator* user_key_comparator,
     Iterator* internal_iter,
-    const SequenceNumber& sequence);
+    SequenceNumber sequence,
+    uint32_t seed);
 
 }  // namespace leveldb
 

db/dbformat.h

@@ -38,6 +38,9 @@ static const int kL0_StopWritesTrigger = 12;
 // space if the same key space is being repeatedly overwritten.
 static const int kMaxMemCompactLevel = 2;
 
+// Approximate gap in bytes between samples of data read during iteration.
+static const int kReadBytesPeriod = 1048576;
+
 }  // namespace config
 
 class InternalKey;

db/version_set.cc

@@ -289,6 +289,51 @@ static bool NewestFirst(FileMetaData* a, FileMetaData* b) {
   return a->number > b->number;
 }
 
+void Version::ForEachOverlapping(Slice user_key, Slice internal_key,
+                                 void* arg,
+                                 bool (*func)(void*, int, FileMetaData*)) {
+  // TODO(sanjay): Change Version::Get() to use this function.
+  const Comparator* ucmp = vset_->icmp_.user_comparator();
+
+  // Search level-0 in order from newest to oldest.
+  std::vector<FileMetaData*> tmp;
+  tmp.reserve(files_[0].size());
+  for (uint32_t i = 0; i < files_[0].size(); i++) {
+    FileMetaData* f = files_[0][i];
+    if (ucmp->Compare(user_key, f->smallest.user_key()) >= 0 &&
+        ucmp->Compare(user_key, f->largest.user_key()) <= 0) {
+      tmp.push_back(f);
+    }
+  }
+  if (!tmp.empty()) {
+    std::sort(tmp.begin(), tmp.end(), NewestFirst);
+    for (uint32_t i = 0; i < tmp.size(); i++) {
+      if (!(*func)(arg, 0, tmp[i])) {
+        return;
+      }
+    }
+  }
+
+  // Search other levels.
+  for (int level = 1; level < config::kNumLevels; level++) {
+    size_t num_files = files_[level].size();
+    if (num_files == 0) continue;
+
+    // Binary search to find earliest index whose largest key >= internal_key.
+    uint32_t index = FindFile(vset_->icmp_, files_[level], internal_key);
+    if (index < num_files) {
+      FileMetaData* f = files_[level][index];
+      if (ucmp->Compare(user_key, f->smallest.user_key()) < 0) {
+        // All of "f" is past any data for user_key
+      } else {
+        if (!(*func)(arg, level, f)) {
+          return;
+        }
+      }
+    }
+  }
+}
+
 Status Version::Get(const ReadOptions& options,
                     const LookupKey& k,
                     std::string* value,
@@ -401,6 +446,44 @@ bool Version::UpdateStats(const GetStats& stats) {
   return false;
 }
 
+bool Version::RecordReadSample(Slice internal_key) {
+  ParsedInternalKey ikey;
+  if (!ParseInternalKey(internal_key, &ikey)) {
+    return false;
+  }
+
+  struct State {
+    GetStats stats;  // Holds first matching file
+    int matches;
+
+    static bool Match(void* arg, int level, FileMetaData* f) {
+      State* state = reinterpret_cast<State*>(arg);
+      state->matches++;
+      if (state->matches == 1) {
+        // Remember first match.
+        state->stats.seek_file = f;
+        state->stats.seek_file_level = level;
+      }
+      // We can stop iterating once we have a second match.
+      return state->matches < 2;
+    }
+  };
+
+  State state;
+  state.matches = 0;
+  ForEachOverlapping(ikey.user_key, internal_key, &state, &State::Match);
+
+  // Must have at least two matches since we want to merge across
+  // files. But what if we have a single file that contains many
+  // overwrites and deletions?  Should we have another mechanism for
+  // finding such files?
+  if (state.matches >= 2) {
+    // 1MB cost is about 1 seek (see comment in Builder::Apply).
+    return UpdateStats(state.stats);
+  }
+  return false;
+}
+
 void Version::Ref() {
   ++refs_;
 }
@@ -435,10 +518,13 @@ int Version::PickLevelForMemTableOutput(
       if (OverlapInLevel(level + 1, &smallest_user_key, &largest_user_key)) {
         break;
       }
-      GetOverlappingInputs(level + 2, &start, &limit, &overlaps);
-      const int64_t sum = TotalFileSize(overlaps);
-      if (sum > kMaxGrandParentOverlapBytes) {
-        break;
+      if (level + 2 < config::kNumLevels) {
+        // Check that file does not overlap too many grandparent bytes.
+        GetOverlappingInputs(level + 2, &start, &limit, &overlaps);
+        const int64_t sum = TotalFileSize(overlaps);
+        if (sum > kMaxGrandParentOverlapBytes) {
+          break;
+        }
       }
       level++;
     }
@@ -452,6 +538,8 @@ void Version::GetOverlappingInputs(
     const InternalKey* begin,
     const InternalKey* end,
     std::vector<FileMetaData*>* inputs) {
+  assert(level >= 0);
+  assert(level < config::kNumLevels);
   inputs->clear();
   Slice user_begin, user_end;
   if (begin != NULL) {

db/version_set.h

@@ -78,6 +78,12 @@ class Version {
   // REQUIRES: lock is held
   bool UpdateStats(const GetStats& stats);
 
+  // Record a sample of bytes read at the specified internal key.
+  // Samples are taken approximately once every config::kReadBytesPeriod
+  // bytes.  Returns true if a new compaction may need to be triggered.
+  // REQUIRES: lock is held
+  bool RecordReadSample(Slice key);
+
   // Reference count management (so Versions do not disappear out from
   // under live iterators)
   void Ref();
@@ -114,6 +120,15 @@ class Version {
   class LevelFileNumIterator;
   Iterator* NewConcatenatingIterator(const ReadOptions&, int level) const;
 
+  // Call func(arg, level, f) for every file that overlaps user_key in
+  // order from newest to oldest.  If an invocation of func returns
+  // false, makes no more calls.
+  //
+  // REQUIRES: user portion of internal_key == user_key.
+  void ForEachOverlapping(Slice user_key, Slice internal_key,
+                          void* arg,
+                          bool (*func)(void*, int, FileMetaData*));
+
   VersionSet* vset_;            // VersionSet to which this Version belongs
   Version* next_;               // Next version in linked list
   Version* prev_;               // Previous version in linked list

include/leveldb/db.h

@@ -14,7 +14,7 @@ namespace leveldb {
 
 // Update Makefile if you change these
 static const int kMajorVersion = 1;
-static const int kMinorVersion = 12;
+static const int kMinorVersion = 13;
 
 struct Options;
 struct ReadOptions;

util/env_posix.cc

@@ -319,8 +319,39 @@ class PosixMmapFile : public WritableFile {
     return Status::OK();
   }
 
-  virtual Status Sync() {
+  Status SyncDirIfManifest() {
+    const char* f = filename_.c_str();
+    const char* sep = strrchr(f, '/');
+    Slice basename;
+    std::string dir;
+    if (sep == NULL) {
+      dir = ".";
+      basename = f;
+    } else {
+      dir = std::string(f, sep - f);
+      basename = sep + 1;
+    }
     Status s;
+    if (basename.starts_with("MANIFEST")) {
+      int fd = open(dir.c_str(), O_RDONLY);
+      if (fd < 0) {
+        s = IOError(dir, errno);
+      } else {
+        if (fsync(fd) < 0) {
+          s = IOError(dir, errno);
+        }
+        close(fd);
+      }
+    }
+    return s;
+  }
+
+  virtual Status Sync() {
+    // Ensure new files referred to by the manifest are in the filesystem.
+    Status s = SyncDirIfManifest();
+    if (!s.ok()) {
+      return s;
+    }
 
     if (pending_sync_) {
       // Some unmapped data was not synced

util/random.h

@@ -16,7 +16,12 @@ class Random {
  private:
   uint32_t seed_;
  public:
-  explicit Random(uint32_t s) : seed_(s & 0x7fffffffu) { }
+  explicit Random(uint32_t s) : seed_(s & 0x7fffffffu) {
+    // Avoid bad seeds.
+    if (seed_ == 0 || seed_ == 2147483647L) {
+      seed_ = 1;
+    }
+  }
   uint32_t Next() {
     static const uint32_t M = 2147483647L;   // 2^31-1
     static const uint64_t A = 16807;  // bits 14, 8, 7, 5, 2, 1, 0