/ src / leveldb / db / db_test.cc
db_test.cc
   1  // Copyright (c) 2011 The LevelDB Authors. All rights reserved.
   2  // Use of this source code is governed by a BSD-style license that can be
   3  // found in the LICENSE file. See the AUTHORS file for names of contributors.
   4  
   5  #include "leveldb/db.h"
   6  
   7  #include <atomic>
   8  #include <string>
   9  
  10  #include "db/db_impl.h"
  11  #include "db/filename.h"
  12  #include "db/version_set.h"
  13  #include "db/write_batch_internal.h"
  14  #include "leveldb/cache.h"
  15  #include "leveldb/env.h"
  16  #include "leveldb/filter_policy.h"
  17  #include "leveldb/table.h"
  18  #include "port/port.h"
  19  #include "port/thread_annotations.h"
  20  #include "util/hash.h"
  21  #include "util/logging.h"
  22  #include "util/mutexlock.h"
  23  #include "util/testharness.h"
  24  #include "util/testutil.h"
  25  
  26  namespace leveldb {
  27  
  28  static std::string RandomString(Random* rnd, int len) {
  29    std::string r;
  30    test::RandomString(rnd, len, &r);
  31    return r;
  32  }
  33  
  34  static std::string RandomKey(Random* rnd) {
  35    int len =
  36        (rnd->OneIn(3) ? 1  // Short sometimes to encourage collisions
  37                       : (rnd->OneIn(100) ? rnd->Skewed(10) : rnd->Uniform(10)));
  38    return test::RandomKey(rnd, len);
  39  }
  40  
  41  namespace {
  42  class AtomicCounter {
  43   public:
  44    AtomicCounter() : count_(0) {}
  45    void Increment() { IncrementBy(1); }
  46    void IncrementBy(int count) LOCKS_EXCLUDED(mu_) {
  47      MutexLock l(&mu_);
  48      count_ += count;
  49    }
  50    int Read() LOCKS_EXCLUDED(mu_) {
  51      MutexLock l(&mu_);
  52      return count_;
  53    }
  54    void Reset() LOCKS_EXCLUDED(mu_) {
  55      MutexLock l(&mu_);
  56      count_ = 0;
  57    }
  58  
  59   private:
  60    port::Mutex mu_;
  61    int count_ GUARDED_BY(mu_);
  62  };
  63  
  64  void DelayMilliseconds(int millis) {
  65    Env::Default()->SleepForMicroseconds(millis * 1000);
  66  }
  67  }  // namespace
  68  
  69  // Test Env to override default Env behavior for testing.
  70  class TestEnv : public EnvWrapper {
  71   public:
  72    explicit TestEnv(Env* base) : EnvWrapper(base), ignore_dot_files_(false) {}
  73  
  74    void SetIgnoreDotFiles(bool ignored) { ignore_dot_files_ = ignored; }
  75  
  76    Status GetChildren(const std::string& dir,
  77                       std::vector<std::string>* result) override {
  78      Status s = target()->GetChildren(dir, result);
  79      if (!s.ok() || !ignore_dot_files_) {
  80        return s;
  81      }
  82  
  83      std::vector<std::string>::iterator it = result->begin();
  84      while (it != result->end()) {
  85        if ((*it == ".") || (*it == "..")) {
  86          it = result->erase(it);
  87        } else {
  88          ++it;
  89        }
  90      }
  91  
  92      return s;
  93    }
  94  
  95   private:
  96    bool ignore_dot_files_;
  97  };
  98  
  99  // Special Env used to delay background operations.
 100  class SpecialEnv : public EnvWrapper {
 101   public:
 102    // sstable/log Sync() calls are blocked while this pointer is non-null.
 103    std::atomic<bool> delay_data_sync_;
 104  
 105    // sstable/log Sync() calls return an error.
 106    std::atomic<bool> data_sync_error_;
 107  
 108    // Simulate no-space errors while this pointer is non-null.
 109    std::atomic<bool> no_space_;
 110  
 111    // Simulate non-writable file system while this pointer is non-null.
 112    std::atomic<bool> non_writable_;
 113  
 114    // Force sync of manifest files to fail while this pointer is non-null.
 115    std::atomic<bool> manifest_sync_error_;
 116  
 117    // Force write to manifest files to fail while this pointer is non-null.
 118    std::atomic<bool> manifest_write_error_;
 119  
 120    bool count_random_reads_;
 121    AtomicCounter random_read_counter_;
 122  
 123    explicit SpecialEnv(Env* base)
 124        : EnvWrapper(base),
 125          delay_data_sync_(false),
 126          data_sync_error_(false),
 127          no_space_(false),
 128          non_writable_(false),
 129          manifest_sync_error_(false),
 130          manifest_write_error_(false),
 131          count_random_reads_(false) {}
 132  
 133    Status NewWritableFile(const std::string& f, WritableFile** r) {
 134      class DataFile : public WritableFile {
 135       private:
 136        SpecialEnv* const env_;
 137        WritableFile* const base_;
 138  
 139       public:
 140        DataFile(SpecialEnv* env, WritableFile* base) : env_(env), base_(base) {}
 141        ~DataFile() { delete base_; }
 142        Status Append(const Slice& data) {
 143          if (env_->no_space_.load(std::memory_order_acquire)) {
 144            // Drop writes on the floor
 145            return Status::OK();
 146          } else {
 147            return base_->Append(data);
 148          }
 149        }
 150        Status Close() { return base_->Close(); }
 151        Status Flush() { return base_->Flush(); }
 152        Status Sync() {
 153          if (env_->data_sync_error_.load(std::memory_order_acquire)) {
 154            return Status::IOError("simulated data sync error");
 155          }
 156          while (env_->delay_data_sync_.load(std::memory_order_acquire)) {
 157            DelayMilliseconds(100);
 158          }
 159          return base_->Sync();
 160        }
 161        std::string GetName() const override { return ""; }
 162      };
 163      class ManifestFile : public WritableFile {
 164       private:
 165        SpecialEnv* env_;
 166        WritableFile* base_;
 167  
 168       public:
 169        ManifestFile(SpecialEnv* env, WritableFile* b) : env_(env), base_(b) {}
 170        ~ManifestFile() { delete base_; }
 171        Status Append(const Slice& data) {
 172          if (env_->manifest_write_error_.load(std::memory_order_acquire)) {
 173            return Status::IOError("simulated writer error");
 174          } else {
 175            return base_->Append(data);
 176          }
 177        }
 178        Status Close() { return base_->Close(); }
 179        Status Flush() { return base_->Flush(); }
 180        Status Sync() {
 181          if (env_->manifest_sync_error_.load(std::memory_order_acquire)) {
 182            return Status::IOError("simulated sync error");
 183          } else {
 184            return base_->Sync();
 185          }
 186        }
 187        std::string GetName() const override { return ""; }
 188      };
 189  
 190      if (non_writable_.load(std::memory_order_acquire)) {
 191        return Status::IOError("simulated write error");
 192      }
 193  
 194      Status s = target()->NewWritableFile(f, r);
 195      if (s.ok()) {
 196        if (strstr(f.c_str(), ".ldb") != nullptr ||
 197            strstr(f.c_str(), ".log") != nullptr) {
 198          *r = new DataFile(this, *r);
 199        } else if (strstr(f.c_str(), "MANIFEST") != nullptr) {
 200          *r = new ManifestFile(this, *r);
 201        }
 202      }
 203      return s;
 204    }
 205  
 206    Status NewRandomAccessFile(const std::string& f, RandomAccessFile** r) {
 207      class CountingFile : public RandomAccessFile {
 208       private:
 209        RandomAccessFile* target_;
 210        AtomicCounter* counter_;
 211  
 212       public:
 213        CountingFile(RandomAccessFile* target, AtomicCounter* counter)
 214            : target_(target), counter_(counter) {}
 215        ~CountingFile() override { delete target_; }
 216        Status Read(uint64_t offset, size_t n, Slice* result,
 217                    char* scratch) const override {
 218          counter_->Increment();
 219          return target_->Read(offset, n, result, scratch);
 220        }
 221        std::string GetName() const override { return ""; }
 222      };
 223  
 224      Status s = target()->NewRandomAccessFile(f, r);
 225      if (s.ok() && count_random_reads_) {
 226        *r = new CountingFile(*r, &random_read_counter_);
 227      }
 228      return s;
 229    }
 230  };
 231  
 232  class DBTest {
 233   public:
 234    std::string dbname_;
 235    SpecialEnv* env_;
 236    DB* db_;
 237  
 238    Options last_options_;
 239  
 240    DBTest() : env_(new SpecialEnv(Env::Default())), option_config_(kDefault) {
 241      filter_policy_ = NewBloomFilterPolicy(10);
 242      dbname_ = test::TmpDir() + "/db_test";
 243      DestroyDB(dbname_, Options());
 244      db_ = nullptr;
 245      Reopen();
 246    }
 247  
 248    ~DBTest() {
 249      delete db_;
 250      DestroyDB(dbname_, Options());
 251      delete env_;
 252      delete filter_policy_;
 253    }
 254  
 255    // Switch to a fresh database with the next option configuration to
 256    // test.  Return false if there are no more configurations to test.
 257    bool ChangeOptions() {
 258      option_config_++;
 259      if (option_config_ >= kEnd) {
 260        return false;
 261      } else {
 262        DestroyAndReopen();
 263        return true;
 264      }
 265    }
 266  
 267    // Return the current option configuration.
 268    Options CurrentOptions() {
 269      Options options;
 270      options.reuse_logs = false;
 271      switch (option_config_) {
 272        case kReuse:
 273          options.reuse_logs = true;
 274          break;
 275        case kFilter:
 276          options.filter_policy = filter_policy_;
 277          break;
 278        case kUncompressed:
 279          options.compression = kNoCompression;
 280          break;
 281        default:
 282          break;
 283      }
 284      return options;
 285    }
 286  
 287    DBImpl* dbfull() { return reinterpret_cast<DBImpl*>(db_); }
 288  
 289    void Reopen(Options* options = nullptr) { ASSERT_OK(TryReopen(options)); }
 290  
 291    void Close() {
 292      delete db_;
 293      db_ = nullptr;
 294    }
 295  
 296    void DestroyAndReopen(Options* options = nullptr) {
 297      delete db_;
 298      db_ = nullptr;
 299      DestroyDB(dbname_, Options());
 300      ASSERT_OK(TryReopen(options));
 301    }
 302  
 303    Status TryReopen(Options* options) {
 304      delete db_;
 305      db_ = nullptr;
 306      Options opts;
 307      if (options != nullptr) {
 308        opts = *options;
 309      } else {
 310        opts = CurrentOptions();
 311        opts.create_if_missing = true;
 312      }
 313      last_options_ = opts;
 314  
 315      return DB::Open(opts, dbname_, &db_);
 316    }
 317  
 318    Status Put(const std::string& k, const std::string& v) {
 319      return db_->Put(WriteOptions(), k, v);
 320    }
 321  
 322    Status Delete(const std::string& k) { return db_->Delete(WriteOptions(), k); }
 323  
 324    std::string Get(const std::string& k, const Snapshot* snapshot = nullptr) {
 325      ReadOptions options;
 326      options.snapshot = snapshot;
 327      std::string result;
 328      Status s = db_->Get(options, k, &result);
 329      if (s.IsNotFound()) {
 330        result = "NOT_FOUND";
 331      } else if (!s.ok()) {
 332        result = s.ToString();
 333      }
 334      return result;
 335    }
 336  
 337    // Return a string that contains all key,value pairs in order,
 338    // formatted like "(k1->v1)(k2->v2)".
 339    std::string Contents() {
 340      std::vector<std::string> forward;
 341      std::string result;
 342      Iterator* iter = db_->NewIterator(ReadOptions());
 343      for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
 344        std::string s = IterStatus(iter);
 345        result.push_back('(');
 346        result.append(s);
 347        result.push_back(')');
 348        forward.push_back(s);
 349      }
 350  
 351      // Check reverse iteration results are the reverse of forward results
 352      size_t matched = 0;
 353      for (iter->SeekToLast(); iter->Valid(); iter->Prev()) {
 354        ASSERT_LT(matched, forward.size());
 355        ASSERT_EQ(IterStatus(iter), forward[forward.size() - matched - 1]);
 356        matched++;
 357      }
 358      ASSERT_EQ(matched, forward.size());
 359  
 360      delete iter;
 361      return result;
 362    }
 363  
 364    std::string AllEntriesFor(const Slice& user_key) {
 365      Iterator* iter = dbfull()->TEST_NewInternalIterator();
 366      InternalKey target(user_key, kMaxSequenceNumber, kTypeValue);
 367      iter->Seek(target.Encode());
 368      std::string result;
 369      if (!iter->status().ok()) {
 370        result = iter->status().ToString();
 371      } else {
 372        result = "[ ";
 373        bool first = true;
 374        while (iter->Valid()) {
 375          ParsedInternalKey ikey;
 376          if (!ParseInternalKey(iter->key(), &ikey)) {
 377            result += "CORRUPTED";
 378          } else {
 379            if (last_options_.comparator->Compare(ikey.user_key, user_key) != 0) {
 380              break;
 381            }
 382            if (!first) {
 383              result += ", ";
 384            }
 385            first = false;
 386            switch (ikey.type) {
 387              case kTypeValue:
 388                result += iter->value().ToString();
 389                break;
 390              case kTypeDeletion:
 391                result += "DEL";
 392                break;
 393            }
 394          }
 395          iter->Next();
 396        }
 397        if (!first) {
 398          result += " ";
 399        }
 400        result += "]";
 401      }
 402      delete iter;
 403      return result;
 404    }
 405  
 406    int NumTableFilesAtLevel(int level) {
 407      std::string property;
 408      ASSERT_TRUE(db_->GetProperty(
 409          "leveldb.num-files-at-level" + NumberToString(level), &property));
 410      return std::stoi(property);
 411    }
 412  
 413    int TotalTableFiles() {
 414      int result = 0;
 415      for (int level = 0; level < config::kNumLevels; level++) {
 416        result += NumTableFilesAtLevel(level);
 417      }
 418      return result;
 419    }
 420  
 421    // Return spread of files per level
 422    std::string FilesPerLevel() {
 423      std::string result;
 424      int last_non_zero_offset = 0;
 425      for (int level = 0; level < config::kNumLevels; level++) {
 426        int f = NumTableFilesAtLevel(level);
 427        char buf[100];
 428        snprintf(buf, sizeof(buf), "%s%d", (level ? "," : ""), f);
 429        result += buf;
 430        if (f > 0) {
 431          last_non_zero_offset = result.size();
 432        }
 433      }
 434      result.resize(last_non_zero_offset);
 435      return result;
 436    }
 437  
 438    int CountFiles() {
 439      std::vector<std::string> files;
 440      env_->GetChildren(dbname_, &files);
 441      return static_cast<int>(files.size());
 442    }
 443  
 444    uint64_t Size(const Slice& start, const Slice& limit) {
 445      Range r(start, limit);
 446      uint64_t size;
 447      db_->GetApproximateSizes(&r, 1, &size);
 448      return size;
 449    }
 450  
 451    void Compact(const Slice& start, const Slice& limit) {
 452      db_->CompactRange(&start, &limit);
 453    }
 454  
 455    // Do n memtable compactions, each of which produces an sstable
 456    // covering the range [small_key,large_key].
 457    void MakeTables(int n, const std::string& small_key,
 458                    const std::string& large_key) {
 459      for (int i = 0; i < n; i++) {
 460        Put(small_key, "begin");
 461        Put(large_key, "end");
 462        dbfull()->TEST_CompactMemTable();
 463      }
 464    }
 465  
 466    // Prevent pushing of new sstables into deeper levels by adding
 467    // tables that cover a specified range to all levels.
 468    void FillLevels(const std::string& smallest, const std::string& largest) {
 469      MakeTables(config::kNumLevels, smallest, largest);
 470    }
 471  
 472    void DumpFileCounts(const char* label) {
 473      fprintf(stderr, "---\n%s:\n", label);
 474      fprintf(
 475          stderr, "maxoverlap: %lld\n",
 476          static_cast<long long>(dbfull()->TEST_MaxNextLevelOverlappingBytes()));
 477      for (int level = 0; level < config::kNumLevels; level++) {
 478        int num = NumTableFilesAtLevel(level);
 479        if (num > 0) {
 480          fprintf(stderr, "  level %3d : %d files\n", level, num);
 481        }
 482      }
 483    }
 484  
 485    std::string DumpSSTableList() {
 486      std::string property;
 487      db_->GetProperty("leveldb.sstables", &property);
 488      return property;
 489    }
 490  
 491    std::string IterStatus(Iterator* iter) {
 492      std::string result;
 493      if (iter->Valid()) {
 494        result = iter->key().ToString() + "->" + iter->value().ToString();
 495      } else {
 496        result = "(invalid)";
 497      }
 498      return result;
 499    }
 500  
 501    bool DeleteAnSSTFile() {
 502      std::vector<std::string> filenames;
 503      ASSERT_OK(env_->GetChildren(dbname_, &filenames));
 504      uint64_t number;
 505      FileType type;
 506      for (size_t i = 0; i < filenames.size(); i++) {
 507        if (ParseFileName(filenames[i], &number, &type) && type == kTableFile) {
 508          ASSERT_OK(env_->DeleteFile(TableFileName(dbname_, number)));
 509          return true;
 510        }
 511      }
 512      return false;
 513    }
 514  
 515    // Returns number of files renamed.
 516    int RenameLDBToSST() {
 517      std::vector<std::string> filenames;
 518      ASSERT_OK(env_->GetChildren(dbname_, &filenames));
 519      uint64_t number;
 520      FileType type;
 521      int files_renamed = 0;
 522      for (size_t i = 0; i < filenames.size(); i++) {
 523        if (ParseFileName(filenames[i], &number, &type) && type == kTableFile) {
 524          const std::string from = TableFileName(dbname_, number);
 525          const std::string to = SSTTableFileName(dbname_, number);
 526          ASSERT_OK(env_->RenameFile(from, to));
 527          files_renamed++;
 528        }
 529      }
 530      return files_renamed;
 531    }
 532  
 533   private:
 534    // Sequence of option configurations to try
 535    enum OptionConfig { kDefault, kReuse, kFilter, kUncompressed, kEnd };
 536  
 537    const FilterPolicy* filter_policy_;
 538    int option_config_;
 539  };
 540  
 541  TEST(DBTest, Empty) {
 542    do {
 543      ASSERT_TRUE(db_ != nullptr);
 544      ASSERT_EQ("NOT_FOUND", Get("foo"));
 545    } while (ChangeOptions());
 546  }
 547  
 548  TEST(DBTest, EmptyKey) {
 549    do {
 550      ASSERT_OK(Put("", "v1"));
 551      ASSERT_EQ("v1", Get(""));
 552      ASSERT_OK(Put("", "v2"));
 553      ASSERT_EQ("v2", Get(""));
 554    } while (ChangeOptions());
 555  }
 556  
 557  TEST(DBTest, EmptyValue) {
 558    do {
 559      ASSERT_OK(Put("key", "v1"));
 560      ASSERT_EQ("v1", Get("key"));
 561      ASSERT_OK(Put("key", ""));
 562      ASSERT_EQ("", Get("key"));
 563      ASSERT_OK(Put("key", "v2"));
 564      ASSERT_EQ("v2", Get("key"));
 565    } while (ChangeOptions());
 566  }
 567  
 568  TEST(DBTest, ReadWrite) {
 569    do {
 570      ASSERT_OK(Put("foo", "v1"));
 571      ASSERT_EQ("v1", Get("foo"));
 572      ASSERT_OK(Put("bar", "v2"));
 573      ASSERT_OK(Put("foo", "v3"));
 574      ASSERT_EQ("v3", Get("foo"));
 575      ASSERT_EQ("v2", Get("bar"));
 576    } while (ChangeOptions());
 577  }
 578  
 579  TEST(DBTest, PutDeleteGet) {
 580    do {
 581      ASSERT_OK(db_->Put(WriteOptions(), "foo", "v1"));
 582      ASSERT_EQ("v1", Get("foo"));
 583      ASSERT_OK(db_->Put(WriteOptions(), "foo", "v2"));
 584      ASSERT_EQ("v2", Get("foo"));
 585      ASSERT_OK(db_->Delete(WriteOptions(), "foo"));
 586      ASSERT_EQ("NOT_FOUND", Get("foo"));
 587    } while (ChangeOptions());
 588  }
 589  
 590  TEST(DBTest, GetFromImmutableLayer) {
 591    do {
 592      Options options = CurrentOptions();
 593      options.env = env_;
 594      options.write_buffer_size = 100000;  // Small write buffer
 595      Reopen(&options);
 596  
 597      ASSERT_OK(Put("foo", "v1"));
 598      ASSERT_EQ("v1", Get("foo"));
 599  
 600      // Block sync calls.
 601      env_->delay_data_sync_.store(true, std::memory_order_release);
 602      Put("k1", std::string(100000, 'x'));  // Fill memtable.
 603      Put("k2", std::string(100000, 'y'));  // Trigger compaction.
 604      ASSERT_EQ("v1", Get("foo"));
 605      // Release sync calls.
 606      env_->delay_data_sync_.store(false, std::memory_order_release);
 607    } while (ChangeOptions());
 608  }
 609  
 610  TEST(DBTest, GetFromVersions) {
 611    do {
 612      ASSERT_OK(Put("foo", "v1"));
 613      dbfull()->TEST_CompactMemTable();
 614      ASSERT_EQ("v1", Get("foo"));
 615    } while (ChangeOptions());
 616  }
 617  
 618  TEST(DBTest, GetMemUsage) {
 619    do {
 620      ASSERT_OK(Put("foo", "v1"));
 621      std::string val;
 622      ASSERT_TRUE(db_->GetProperty("leveldb.approximate-memory-usage", &val));
 623      int mem_usage = std::stoi(val);
 624      ASSERT_GT(mem_usage, 0);
 625      ASSERT_LT(mem_usage, 5 * 1024 * 1024);
 626    } while (ChangeOptions());
 627  }
 628  
 629  TEST(DBTest, GetSnapshot) {
 630    do {
 631      // Try with both a short key and a long key
 632      for (int i = 0; i < 2; i++) {
 633        std::string key = (i == 0) ? std::string("foo") : std::string(200, 'x');
 634        ASSERT_OK(Put(key, "v1"));
 635        const Snapshot* s1 = db_->GetSnapshot();
 636        ASSERT_OK(Put(key, "v2"));
 637        ASSERT_EQ("v2", Get(key));
 638        ASSERT_EQ("v1", Get(key, s1));
 639        dbfull()->TEST_CompactMemTable();
 640        ASSERT_EQ("v2", Get(key));
 641        ASSERT_EQ("v1", Get(key, s1));
 642        db_->ReleaseSnapshot(s1);
 643      }
 644    } while (ChangeOptions());
 645  }
 646  
 647  TEST(DBTest, GetIdenticalSnapshots) {
 648    do {
 649      // Try with both a short key and a long key
 650      for (int i = 0; i < 2; i++) {
 651        std::string key = (i == 0) ? std::string("foo") : std::string(200, 'x');
 652        ASSERT_OK(Put(key, "v1"));
 653        const Snapshot* s1 = db_->GetSnapshot();
 654        const Snapshot* s2 = db_->GetSnapshot();
 655        const Snapshot* s3 = db_->GetSnapshot();
 656        ASSERT_OK(Put(key, "v2"));
 657        ASSERT_EQ("v2", Get(key));
 658        ASSERT_EQ("v1", Get(key, s1));
 659        ASSERT_EQ("v1", Get(key, s2));
 660        ASSERT_EQ("v1", Get(key, s3));
 661        db_->ReleaseSnapshot(s1);
 662        dbfull()->TEST_CompactMemTable();
 663        ASSERT_EQ("v2", Get(key));
 664        ASSERT_EQ("v1", Get(key, s2));
 665        db_->ReleaseSnapshot(s2);
 666        ASSERT_EQ("v1", Get(key, s3));
 667        db_->ReleaseSnapshot(s3);
 668      }
 669    } while (ChangeOptions());
 670  }
 671  
 672  TEST(DBTest, IterateOverEmptySnapshot) {
 673    do {
 674      const Snapshot* snapshot = db_->GetSnapshot();
 675      ReadOptions read_options;
 676      read_options.snapshot = snapshot;
 677      ASSERT_OK(Put("foo", "v1"));
 678      ASSERT_OK(Put("foo", "v2"));
 679  
 680      Iterator* iterator1 = db_->NewIterator(read_options);
 681      iterator1->SeekToFirst();
 682      ASSERT_TRUE(!iterator1->Valid());
 683      delete iterator1;
 684  
 685      dbfull()->TEST_CompactMemTable();
 686  
 687      Iterator* iterator2 = db_->NewIterator(read_options);
 688      iterator2->SeekToFirst();
 689      ASSERT_TRUE(!iterator2->Valid());
 690      delete iterator2;
 691  
 692      db_->ReleaseSnapshot(snapshot);
 693    } while (ChangeOptions());
 694  }
 695  
 696  TEST(DBTest, GetLevel0Ordering) {
 697    do {
 698      // Check that we process level-0 files in correct order.  The code
 699      // below generates two level-0 files where the earlier one comes
 700      // before the later one in the level-0 file list since the earlier
 701      // one has a smaller "smallest" key.
 702      ASSERT_OK(Put("bar", "b"));
 703      ASSERT_OK(Put("foo", "v1"));
 704      dbfull()->TEST_CompactMemTable();
 705      ASSERT_OK(Put("foo", "v2"));
 706      dbfull()->TEST_CompactMemTable();
 707      ASSERT_EQ("v2", Get("foo"));
 708    } while (ChangeOptions());
 709  }
 710  
 711  TEST(DBTest, GetOrderedByLevels) {
 712    do {
 713      ASSERT_OK(Put("foo", "v1"));
 714      Compact("a", "z");
 715      ASSERT_EQ("v1", Get("foo"));
 716      ASSERT_OK(Put("foo", "v2"));
 717      ASSERT_EQ("v2", Get("foo"));
 718      dbfull()->TEST_CompactMemTable();
 719      ASSERT_EQ("v2", Get("foo"));
 720    } while (ChangeOptions());
 721  }
 722  
 723  TEST(DBTest, GetPicksCorrectFile) {
 724    do {
 725      // Arrange to have multiple files in a non-level-0 level.
 726      ASSERT_OK(Put("a", "va"));
 727      Compact("a", "b");
 728      ASSERT_OK(Put("x", "vx"));
 729      Compact("x", "y");
 730      ASSERT_OK(Put("f", "vf"));
 731      Compact("f", "g");
 732      ASSERT_EQ("va", Get("a"));
 733      ASSERT_EQ("vf", Get("f"));
 734      ASSERT_EQ("vx", Get("x"));
 735    } while (ChangeOptions());
 736  }
 737  
 738  TEST(DBTest, GetEncountersEmptyLevel) {
 739    do {
 740      // Arrange for the following to happen:
 741      //   * sstable A in level 0
 742      //   * nothing in level 1
 743      //   * sstable B in level 2
 744      // Then do enough Get() calls to arrange for an automatic compaction
 745      // of sstable A.  A bug would cause the compaction to be marked as
 746      // occurring at level 1 (instead of the correct level 0).
 747  
 748      // Step 1: First place sstables in levels 0 and 2
 749      int compaction_count = 0;
 750      while (NumTableFilesAtLevel(0) == 0 || NumTableFilesAtLevel(2) == 0) {
 751        ASSERT_LE(compaction_count, 100) << "could not fill levels 0 and 2";
 752        compaction_count++;
 753        Put("a", "begin");
 754        Put("z", "end");
 755        dbfull()->TEST_CompactMemTable();
 756      }
 757  
 758      // Step 2: clear level 1 if necessary.
 759      dbfull()->TEST_CompactRange(1, nullptr, nullptr);
 760      ASSERT_EQ(NumTableFilesAtLevel(0), 1);
 761      ASSERT_EQ(NumTableFilesAtLevel(1), 0);
 762      ASSERT_EQ(NumTableFilesAtLevel(2), 1);
 763  
 764      // Step 3: read a bunch of times
 765      for (int i = 0; i < 1000; i++) {
 766        ASSERT_EQ("NOT_FOUND", Get("missing"));
 767      }
 768  
 769      // Step 4: Wait for compaction to finish
 770      DelayMilliseconds(1000);
 771  
 772      ASSERT_EQ(NumTableFilesAtLevel(0), 0);
 773    } while (ChangeOptions());
 774  }
 775  
 776  TEST(DBTest, IterEmpty) {
 777    Iterator* iter = db_->NewIterator(ReadOptions());
 778  
 779    iter->SeekToFirst();
 780    ASSERT_EQ(IterStatus(iter), "(invalid)");
 781  
 782    iter->SeekToLast();
 783    ASSERT_EQ(IterStatus(iter), "(invalid)");
 784  
 785    iter->Seek("foo");
 786    ASSERT_EQ(IterStatus(iter), "(invalid)");
 787  
 788    delete iter;
 789  }
 790  
 791  TEST(DBTest, IterSingle) {
 792    ASSERT_OK(Put("a", "va"));
 793    Iterator* iter = db_->NewIterator(ReadOptions());
 794  
 795    iter->SeekToFirst();
 796    ASSERT_EQ(IterStatus(iter), "a->va");
 797    iter->Next();
 798    ASSERT_EQ(IterStatus(iter), "(invalid)");
 799    iter->SeekToFirst();
 800    ASSERT_EQ(IterStatus(iter), "a->va");
 801    iter->Prev();
 802    ASSERT_EQ(IterStatus(iter), "(invalid)");
 803  
 804    iter->SeekToLast();
 805    ASSERT_EQ(IterStatus(iter), "a->va");
 806    iter->Next();
 807    ASSERT_EQ(IterStatus(iter), "(invalid)");
 808    iter->SeekToLast();
 809    ASSERT_EQ(IterStatus(iter), "a->va");
 810    iter->Prev();
 811    ASSERT_EQ(IterStatus(iter), "(invalid)");
 812  
 813    iter->Seek("");
 814    ASSERT_EQ(IterStatus(iter), "a->va");
 815    iter->Next();
 816    ASSERT_EQ(IterStatus(iter), "(invalid)");
 817  
 818    iter->Seek("a");
 819    ASSERT_EQ(IterStatus(iter), "a->va");
 820    iter->Next();
 821    ASSERT_EQ(IterStatus(iter), "(invalid)");
 822  
 823    iter->Seek("b");
 824    ASSERT_EQ(IterStatus(iter), "(invalid)");
 825  
 826    delete iter;
 827  }
 828  
 829  TEST(DBTest, IterMulti) {
 830    ASSERT_OK(Put("a", "va"));
 831    ASSERT_OK(Put("b", "vb"));
 832    ASSERT_OK(Put("c", "vc"));
 833    Iterator* iter = db_->NewIterator(ReadOptions());
 834  
 835    iter->SeekToFirst();
 836    ASSERT_EQ(IterStatus(iter), "a->va");
 837    iter->Next();
 838    ASSERT_EQ(IterStatus(iter), "b->vb");
 839    iter->Next();
 840    ASSERT_EQ(IterStatus(iter), "c->vc");
 841    iter->Next();
 842    ASSERT_EQ(IterStatus(iter), "(invalid)");
 843    iter->SeekToFirst();
 844    ASSERT_EQ(IterStatus(iter), "a->va");
 845    iter->Prev();
 846    ASSERT_EQ(IterStatus(iter), "(invalid)");
 847  
 848    iter->SeekToLast();
 849    ASSERT_EQ(IterStatus(iter), "c->vc");
 850    iter->Prev();
 851    ASSERT_EQ(IterStatus(iter), "b->vb");
 852    iter->Prev();
 853    ASSERT_EQ(IterStatus(iter), "a->va");
 854    iter->Prev();
 855    ASSERT_EQ(IterStatus(iter), "(invalid)");
 856    iter->SeekToLast();
 857    ASSERT_EQ(IterStatus(iter), "c->vc");
 858    iter->Next();
 859    ASSERT_EQ(IterStatus(iter), "(invalid)");
 860  
 861    iter->Seek("");
 862    ASSERT_EQ(IterStatus(iter), "a->va");
 863    iter->Seek("a");
 864    ASSERT_EQ(IterStatus(iter), "a->va");
 865    iter->Seek("ax");
 866    ASSERT_EQ(IterStatus(iter), "b->vb");
 867    iter->Seek("b");
 868    ASSERT_EQ(IterStatus(iter), "b->vb");
 869    iter->Seek("z");
 870    ASSERT_EQ(IterStatus(iter), "(invalid)");
 871  
 872    // Switch from reverse to forward
 873    iter->SeekToLast();
 874    iter->Prev();
 875    iter->Prev();
 876    iter->Next();
 877    ASSERT_EQ(IterStatus(iter), "b->vb");
 878  
 879    // Switch from forward to reverse
 880    iter->SeekToFirst();
 881    iter->Next();
 882    iter->Next();
 883    iter->Prev();
 884    ASSERT_EQ(IterStatus(iter), "b->vb");
 885  
 886    // Make sure iter stays at snapshot
 887    ASSERT_OK(Put("a", "va2"));
 888    ASSERT_OK(Put("a2", "va3"));
 889    ASSERT_OK(Put("b", "vb2"));
 890    ASSERT_OK(Put("c", "vc2"));
 891    ASSERT_OK(Delete("b"));
 892    iter->SeekToFirst();
 893    ASSERT_EQ(IterStatus(iter), "a->va");
 894    iter->Next();
 895    ASSERT_EQ(IterStatus(iter), "b->vb");
 896    iter->Next();
 897    ASSERT_EQ(IterStatus(iter), "c->vc");
 898    iter->Next();
 899    ASSERT_EQ(IterStatus(iter), "(invalid)");
 900    iter->SeekToLast();
 901    ASSERT_EQ(IterStatus(iter), "c->vc");
 902    iter->Prev();
 903    ASSERT_EQ(IterStatus(iter), "b->vb");
 904    iter->Prev();
 905    ASSERT_EQ(IterStatus(iter), "a->va");
 906    iter->Prev();
 907    ASSERT_EQ(IterStatus(iter), "(invalid)");
 908  
 909    delete iter;
 910  }
 911  
 912  TEST(DBTest, IterSmallAndLargeMix) {
 913    ASSERT_OK(Put("a", "va"));
 914    ASSERT_OK(Put("b", std::string(100000, 'b')));
 915    ASSERT_OK(Put("c", "vc"));
 916    ASSERT_OK(Put("d", std::string(100000, 'd')));
 917    ASSERT_OK(Put("e", std::string(100000, 'e')));
 918  
 919    Iterator* iter = db_->NewIterator(ReadOptions());
 920  
 921    iter->SeekToFirst();
 922    ASSERT_EQ(IterStatus(iter), "a->va");
 923    iter->Next();
 924    ASSERT_EQ(IterStatus(iter), "b->" + std::string(100000, 'b'));
 925    iter->Next();
 926    ASSERT_EQ(IterStatus(iter), "c->vc");
 927    iter->Next();
 928    ASSERT_EQ(IterStatus(iter), "d->" + std::string(100000, 'd'));
 929    iter->Next();
 930    ASSERT_EQ(IterStatus(iter), "e->" + std::string(100000, 'e'));
 931    iter->Next();
 932    ASSERT_EQ(IterStatus(iter), "(invalid)");
 933  
 934    iter->SeekToLast();
 935    ASSERT_EQ(IterStatus(iter), "e->" + std::string(100000, 'e'));
 936    iter->Prev();
 937    ASSERT_EQ(IterStatus(iter), "d->" + std::string(100000, 'd'));
 938    iter->Prev();
 939    ASSERT_EQ(IterStatus(iter), "c->vc");
 940    iter->Prev();
 941    ASSERT_EQ(IterStatus(iter), "b->" + std::string(100000, 'b'));
 942    iter->Prev();
 943    ASSERT_EQ(IterStatus(iter), "a->va");
 944    iter->Prev();
 945    ASSERT_EQ(IterStatus(iter), "(invalid)");
 946  
 947    delete iter;
 948  }
 949  
 950  TEST(DBTest, IterMultiWithDelete) {
 951    do {
 952      ASSERT_OK(Put("a", "va"));
 953      ASSERT_OK(Put("b", "vb"));
 954      ASSERT_OK(Put("c", "vc"));
 955      ASSERT_OK(Delete("b"));
 956      ASSERT_EQ("NOT_FOUND", Get("b"));
 957  
 958      Iterator* iter = db_->NewIterator(ReadOptions());
 959      iter->Seek("c");
 960      ASSERT_EQ(IterStatus(iter), "c->vc");
 961      iter->Prev();
 962      ASSERT_EQ(IterStatus(iter), "a->va");
 963      delete iter;
 964    } while (ChangeOptions());
 965  }
 966  
 967  TEST(DBTest, Recover) {
 968    do {
 969      ASSERT_OK(Put("foo", "v1"));
 970      ASSERT_OK(Put("baz", "v5"));
 971  
 972      Reopen();
 973      ASSERT_EQ("v1", Get("foo"));
 974  
 975      ASSERT_EQ("v1", Get("foo"));
 976      ASSERT_EQ("v5", Get("baz"));
 977      ASSERT_OK(Put("bar", "v2"));
 978      ASSERT_OK(Put("foo", "v3"));
 979  
 980      Reopen();
 981      ASSERT_EQ("v3", Get("foo"));
 982      ASSERT_OK(Put("foo", "v4"));
 983      ASSERT_EQ("v4", Get("foo"));
 984      ASSERT_EQ("v2", Get("bar"));
 985      ASSERT_EQ("v5", Get("baz"));
 986    } while (ChangeOptions());
 987  }
 988  
 989  TEST(DBTest, RecoveryWithEmptyLog) {
 990    do {
 991      ASSERT_OK(Put("foo", "v1"));
 992      ASSERT_OK(Put("foo", "v2"));
 993      Reopen();
 994      Reopen();
 995      ASSERT_OK(Put("foo", "v3"));
 996      Reopen();
 997      ASSERT_EQ("v3", Get("foo"));
 998    } while (ChangeOptions());
 999  }
1000  
1001  // Check that writes done during a memtable compaction are recovered
1002  // if the database is shutdown during the memtable compaction.
1003  TEST(DBTest, RecoverDuringMemtableCompaction) {
1004    do {
1005      Options options = CurrentOptions();
1006      options.env = env_;
1007      options.write_buffer_size = 1000000;
1008      Reopen(&options);
1009  
1010      // Trigger a long memtable compaction and reopen the database during it
1011      ASSERT_OK(Put("foo", "v1"));                         // Goes to 1st log file
1012      ASSERT_OK(Put("big1", std::string(10000000, 'x')));  // Fills memtable
1013      ASSERT_OK(Put("big2", std::string(1000, 'y')));      // Triggers compaction
1014      ASSERT_OK(Put("bar", "v2"));                         // Goes to new log file
1015  
1016      Reopen(&options);
1017      ASSERT_EQ("v1", Get("foo"));
1018      ASSERT_EQ("v2", Get("bar"));
1019      ASSERT_EQ(std::string(10000000, 'x'), Get("big1"));
1020      ASSERT_EQ(std::string(1000, 'y'), Get("big2"));
1021    } while (ChangeOptions());
1022  }
1023  
1024  static std::string Key(int i) {
1025    char buf[100];
1026    snprintf(buf, sizeof(buf), "key%06d", i);
1027    return std::string(buf);
1028  }
1029  
1030  TEST(DBTest, MinorCompactionsHappen) {
1031    Options options = CurrentOptions();
1032    options.write_buffer_size = 10000;
1033    Reopen(&options);
1034  
1035    const int N = 500;
1036  
1037    int starting_num_tables = TotalTableFiles();
1038    for (int i = 0; i < N; i++) {
1039      ASSERT_OK(Put(Key(i), Key(i) + std::string(1000, 'v')));
1040    }
1041    int ending_num_tables = TotalTableFiles();
1042    ASSERT_GT(ending_num_tables, starting_num_tables);
1043  
1044    for (int i = 0; i < N; i++) {
1045      ASSERT_EQ(Key(i) + std::string(1000, 'v'), Get(Key(i)));
1046    }
1047  
1048    Reopen();
1049  
1050    for (int i = 0; i < N; i++) {
1051      ASSERT_EQ(Key(i) + std::string(1000, 'v'), Get(Key(i)));
1052    }
1053  }
1054  
1055  TEST(DBTest, RecoverWithLargeLog) {
1056    {
1057      Options options = CurrentOptions();
1058      Reopen(&options);
1059      ASSERT_OK(Put("big1", std::string(200000, '1')));
1060      ASSERT_OK(Put("big2", std::string(200000, '2')));
1061      ASSERT_OK(Put("small3", std::string(10, '3')));
1062      ASSERT_OK(Put("small4", std::string(10, '4')));
1063      ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1064    }
1065  
1066    // Make sure that if we re-open with a small write buffer size that
1067    // we flush table files in the middle of a large log file.
1068    Options options = CurrentOptions();
1069    options.write_buffer_size = 100000;
1070    Reopen(&options);
1071    ASSERT_EQ(NumTableFilesAtLevel(0), 3);
1072    ASSERT_EQ(std::string(200000, '1'), Get("big1"));
1073    ASSERT_EQ(std::string(200000, '2'), Get("big2"));
1074    ASSERT_EQ(std::string(10, '3'), Get("small3"));
1075    ASSERT_EQ(std::string(10, '4'), Get("small4"));
1076    ASSERT_GT(NumTableFilesAtLevel(0), 1);
1077  }
1078  
1079  TEST(DBTest, CompactionsGenerateMultipleFiles) {
1080    Options options = CurrentOptions();
1081    options.write_buffer_size = 100000000;  // Large write buffer
1082    Reopen(&options);
1083  
1084    Random rnd(301);
1085  
1086    // Write 8MB (80 values, each 100K)
1087    ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1088    std::vector<std::string> values;
1089    for (int i = 0; i < 80; i++) {
1090      values.push_back(RandomString(&rnd, 100000));
1091      ASSERT_OK(Put(Key(i), values[i]));
1092    }
1093  
1094    // Reopening moves updates to level-0
1095    Reopen(&options);
1096    dbfull()->TEST_CompactRange(0, nullptr, nullptr);
1097  
1098    ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1099    ASSERT_GT(NumTableFilesAtLevel(1), 1);
1100    for (int i = 0; i < 80; i++) {
1101      ASSERT_EQ(Get(Key(i)), values[i]);
1102    }
1103  }
1104  
1105  TEST(DBTest, RepeatedWritesToSameKey) {
1106    Options options = CurrentOptions();
1107    options.env = env_;
1108    options.write_buffer_size = 100000;  // Small write buffer
1109    Reopen(&options);
1110  
1111    // We must have at most one file per level except for level-0,
1112    // which may have up to kL0_StopWritesTrigger files.
1113    const int kMaxFiles = config::kNumLevels + config::kL0_StopWritesTrigger;
1114  
1115    Random rnd(301);
1116    std::string value = RandomString(&rnd, 2 * options.write_buffer_size);
1117    for (int i = 0; i < 5 * kMaxFiles; i++) {
1118      Put("key", value);
1119      ASSERT_LE(TotalTableFiles(), kMaxFiles);
1120      fprintf(stderr, "after %d: %d files\n", i + 1, TotalTableFiles());
1121    }
1122  }
1123  
1124  TEST(DBTest, SparseMerge) {
1125    Options options = CurrentOptions();
1126    options.compression = kNoCompression;
1127    Reopen(&options);
1128  
1129    FillLevels("A", "Z");
1130  
1131    // Suppose there is:
1132    //    small amount of data with prefix A
1133    //    large amount of data with prefix B
1134    //    small amount of data with prefix C
1135    // and that recent updates have made small changes to all three prefixes.
1136    // Check that we do not do a compaction that merges all of B in one shot.
1137    const std::string value(1000, 'x');
1138    Put("A", "va");
1139    // Write approximately 100MB of "B" values
1140    for (int i = 0; i < 100000; i++) {
1141      char key[100];
1142      snprintf(key, sizeof(key), "B%010d", i);
1143      Put(key, value);
1144    }
1145    Put("C", "vc");
1146    dbfull()->TEST_CompactMemTable();
1147    dbfull()->TEST_CompactRange(0, nullptr, nullptr);
1148  
1149    // Make sparse update
1150    Put("A", "va2");
1151    Put("B100", "bvalue2");
1152    Put("C", "vc2");
1153    dbfull()->TEST_CompactMemTable();
1154  
1155    // Compactions should not cause us to create a situation where
1156    // a file overlaps too much data at the next level.
1157    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(), 20 * 1048576);
1158    dbfull()->TEST_CompactRange(0, nullptr, nullptr);
1159    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(), 20 * 1048576);
1160    dbfull()->TEST_CompactRange(1, nullptr, nullptr);
1161    ASSERT_LE(dbfull()->TEST_MaxNextLevelOverlappingBytes(), 20 * 1048576);
1162  }
1163  
1164  static bool Between(uint64_t val, uint64_t low, uint64_t high) {
1165    bool result = (val >= low) && (val <= high);
1166    if (!result) {
1167      fprintf(stderr, "Value %llu is not in range [%llu, %llu]\n",
1168              (unsigned long long)(val), (unsigned long long)(low),
1169              (unsigned long long)(high));
1170    }
1171    return result;
1172  }
1173  
1174  TEST(DBTest, ApproximateSizes) {
1175    do {
1176      Options options = CurrentOptions();
1177      options.write_buffer_size = 100000000;  // Large write buffer
1178      options.compression = kNoCompression;
1179      DestroyAndReopen();
1180  
1181      ASSERT_TRUE(Between(Size("", "xyz"), 0, 0));
1182      Reopen(&options);
1183      ASSERT_TRUE(Between(Size("", "xyz"), 0, 0));
1184  
1185      // Write 8MB (80 values, each 100K)
1186      ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1187      const int N = 80;
1188      static const int S1 = 100000;
1189      static const int S2 = 105000;  // Allow some expansion from metadata
1190      Random rnd(301);
1191      for (int i = 0; i < N; i++) {
1192        ASSERT_OK(Put(Key(i), RandomString(&rnd, S1)));
1193      }
1194  
1195      // 0 because GetApproximateSizes() does not account for memtable space
1196      ASSERT_TRUE(Between(Size("", Key(50)), 0, 0));
1197  
1198      if (options.reuse_logs) {
1199        // Recovery will reuse memtable, and GetApproximateSizes() does not
1200        // account for memtable usage;
1201        Reopen(&options);
1202        ASSERT_TRUE(Between(Size("", Key(50)), 0, 0));
1203        continue;
1204      }
1205  
1206      // Check sizes across recovery by reopening a few times
1207      for (int run = 0; run < 3; run++) {
1208        Reopen(&options);
1209  
1210        for (int compact_start = 0; compact_start < N; compact_start += 10) {
1211          for (int i = 0; i < N; i += 10) {
1212            ASSERT_TRUE(Between(Size("", Key(i)), S1 * i, S2 * i));
1213            ASSERT_TRUE(Between(Size("", Key(i) + ".suffix"), S1 * (i + 1),
1214                                S2 * (i + 1)));
1215            ASSERT_TRUE(Between(Size(Key(i), Key(i + 10)), S1 * 10, S2 * 10));
1216          }
1217          ASSERT_TRUE(Between(Size("", Key(50)), S1 * 50, S2 * 50));
1218          ASSERT_TRUE(Between(Size("", Key(50) + ".suffix"), S1 * 50, S2 * 50));
1219  
1220          std::string cstart_str = Key(compact_start);
1221          std::string cend_str = Key(compact_start + 9);
1222          Slice cstart = cstart_str;
1223          Slice cend = cend_str;
1224          dbfull()->TEST_CompactRange(0, &cstart, &cend);
1225        }
1226  
1227        ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1228        ASSERT_GT(NumTableFilesAtLevel(1), 0);
1229      }
1230    } while (ChangeOptions());
1231  }
1232  
1233  TEST(DBTest, ApproximateSizes_MixOfSmallAndLarge) {
1234    do {
1235      Options options = CurrentOptions();
1236      options.compression = kNoCompression;
1237      Reopen();
1238  
1239      Random rnd(301);
1240      std::string big1 = RandomString(&rnd, 100000);
1241      ASSERT_OK(Put(Key(0), RandomString(&rnd, 10000)));
1242      ASSERT_OK(Put(Key(1), RandomString(&rnd, 10000)));
1243      ASSERT_OK(Put(Key(2), big1));
1244      ASSERT_OK(Put(Key(3), RandomString(&rnd, 10000)));
1245      ASSERT_OK(Put(Key(4), big1));
1246      ASSERT_OK(Put(Key(5), RandomString(&rnd, 10000)));
1247      ASSERT_OK(Put(Key(6), RandomString(&rnd, 300000)));
1248      ASSERT_OK(Put(Key(7), RandomString(&rnd, 10000)));
1249  
1250      if (options.reuse_logs) {
1251        // Need to force a memtable compaction since recovery does not do so.
1252        ASSERT_OK(dbfull()->TEST_CompactMemTable());
1253      }
1254  
1255      // Check sizes across recovery by reopening a few times
1256      for (int run = 0; run < 3; run++) {
1257        Reopen(&options);
1258  
1259        ASSERT_TRUE(Between(Size("", Key(0)), 0, 0));
1260        ASSERT_TRUE(Between(Size("", Key(1)), 10000, 11000));
1261        ASSERT_TRUE(Between(Size("", Key(2)), 20000, 21000));
1262        ASSERT_TRUE(Between(Size("", Key(3)), 120000, 121000));
1263        ASSERT_TRUE(Between(Size("", Key(4)), 130000, 131000));
1264        ASSERT_TRUE(Between(Size("", Key(5)), 230000, 231000));
1265        ASSERT_TRUE(Between(Size("", Key(6)), 240000, 241000));
1266        ASSERT_TRUE(Between(Size("", Key(7)), 540000, 541000));
1267        ASSERT_TRUE(Between(Size("", Key(8)), 550000, 560000));
1268  
1269        ASSERT_TRUE(Between(Size(Key(3), Key(5)), 110000, 111000));
1270  
1271        dbfull()->TEST_CompactRange(0, nullptr, nullptr);
1272      }
1273    } while (ChangeOptions());
1274  }
1275  
1276  TEST(DBTest, IteratorPinsRef) {
1277    Put("foo", "hello");
1278  
1279    // Get iterator that will yield the current contents of the DB.
1280    Iterator* iter = db_->NewIterator(ReadOptions());
1281  
1282    // Write to force compactions
1283    Put("foo", "newvalue1");
1284    for (int i = 0; i < 100; i++) {
1285      ASSERT_OK(Put(Key(i), Key(i) + std::string(100000, 'v')));  // 100K values
1286    }
1287    Put("foo", "newvalue2");
1288  
1289    iter->SeekToFirst();
1290    ASSERT_TRUE(iter->Valid());
1291    ASSERT_EQ("foo", iter->key().ToString());
1292    ASSERT_EQ("hello", iter->value().ToString());
1293    iter->Next();
1294    ASSERT_TRUE(!iter->Valid());
1295    delete iter;
1296  }
1297  
1298  TEST(DBTest, Snapshot) {
1299    do {
1300      Put("foo", "v1");
1301      const Snapshot* s1 = db_->GetSnapshot();
1302      Put("foo", "v2");
1303      const Snapshot* s2 = db_->GetSnapshot();
1304      Put("foo", "v3");
1305      const Snapshot* s3 = db_->GetSnapshot();
1306  
1307      Put("foo", "v4");
1308      ASSERT_EQ("v1", Get("foo", s1));
1309      ASSERT_EQ("v2", Get("foo", s2));
1310      ASSERT_EQ("v3", Get("foo", s3));
1311      ASSERT_EQ("v4", Get("foo"));
1312  
1313      db_->ReleaseSnapshot(s3);
1314      ASSERT_EQ("v1", Get("foo", s1));
1315      ASSERT_EQ("v2", Get("foo", s2));
1316      ASSERT_EQ("v4", Get("foo"));
1317  
1318      db_->ReleaseSnapshot(s1);
1319      ASSERT_EQ("v2", Get("foo", s2));
1320      ASSERT_EQ("v4", Get("foo"));
1321  
1322      db_->ReleaseSnapshot(s2);
1323      ASSERT_EQ("v4", Get("foo"));
1324    } while (ChangeOptions());
1325  }
1326  
1327  TEST(DBTest, HiddenValuesAreRemoved) {
1328    do {
1329      Random rnd(301);
1330      FillLevels("a", "z");
1331  
1332      std::string big = RandomString(&rnd, 50000);
1333      Put("foo", big);
1334      Put("pastfoo", "v");
1335      const Snapshot* snapshot = db_->GetSnapshot();
1336      Put("foo", "tiny");
1337      Put("pastfoo2", "v2");  // Advance sequence number one more
1338  
1339      ASSERT_OK(dbfull()->TEST_CompactMemTable());
1340      ASSERT_GT(NumTableFilesAtLevel(0), 0);
1341  
1342      ASSERT_EQ(big, Get("foo", snapshot));
1343      ASSERT_TRUE(Between(Size("", "pastfoo"), 50000, 60000));
1344      db_->ReleaseSnapshot(snapshot);
1345      ASSERT_EQ(AllEntriesFor("foo"), "[ tiny, " + big + " ]");
1346      Slice x("x");
1347      dbfull()->TEST_CompactRange(0, nullptr, &x);
1348      ASSERT_EQ(AllEntriesFor("foo"), "[ tiny ]");
1349      ASSERT_EQ(NumTableFilesAtLevel(0), 0);
1350      ASSERT_GE(NumTableFilesAtLevel(1), 1);
1351      dbfull()->TEST_CompactRange(1, nullptr, &x);
1352      ASSERT_EQ(AllEntriesFor("foo"), "[ tiny ]");
1353  
1354      ASSERT_TRUE(Between(Size("", "pastfoo"), 0, 1000));
1355    } while (ChangeOptions());
1356  }
1357  
1358  TEST(DBTest, DeletionMarkers1) {
1359    Put("foo", "v1");
1360    ASSERT_OK(dbfull()->TEST_CompactMemTable());
1361    const int last = config::kMaxMemCompactLevel;
1362    ASSERT_EQ(NumTableFilesAtLevel(last), 1);  // foo => v1 is now in last level
1363  
1364    // Place a table at level last-1 to prevent merging with preceding mutation
1365    Put("a", "begin");
1366    Put("z", "end");
1367    dbfull()->TEST_CompactMemTable();
1368    ASSERT_EQ(NumTableFilesAtLevel(last), 1);
1369    ASSERT_EQ(NumTableFilesAtLevel(last - 1), 1);
1370  
1371    Delete("foo");
1372    Put("foo", "v2");
1373    ASSERT_EQ(AllEntriesFor("foo"), "[ v2, DEL, v1 ]");
1374    ASSERT_OK(dbfull()->TEST_CompactMemTable());  // Moves to level last-2
1375    ASSERT_EQ(AllEntriesFor("foo"), "[ v2, DEL, v1 ]");
1376    Slice z("z");
1377    dbfull()->TEST_CompactRange(last - 2, nullptr, &z);
1378    // DEL eliminated, but v1 remains because we aren't compacting that level
1379    // (DEL can be eliminated because v2 hides v1).
1380    ASSERT_EQ(AllEntriesFor("foo"), "[ v2, v1 ]");
1381    dbfull()->TEST_CompactRange(last - 1, nullptr, nullptr);
1382    // Merging last-1 w/ last, so we are the base level for "foo", so
1383    // DEL is removed.  (as is v1).
1384    ASSERT_EQ(AllEntriesFor("foo"), "[ v2 ]");
1385  }
1386  
1387  TEST(DBTest, DeletionMarkers2) {
1388    Put("foo", "v1");
1389    ASSERT_OK(dbfull()->TEST_CompactMemTable());
1390    const int last = config::kMaxMemCompactLevel;
1391    ASSERT_EQ(NumTableFilesAtLevel(last), 1);  // foo => v1 is now in last level
1392  
1393    // Place a table at level last-1 to prevent merging with preceding mutation
1394    Put("a", "begin");
1395    Put("z", "end");
1396    dbfull()->TEST_CompactMemTable();
1397    ASSERT_EQ(NumTableFilesAtLevel(last), 1);
1398    ASSERT_EQ(NumTableFilesAtLevel(last - 1), 1);
1399  
1400    Delete("foo");
1401    ASSERT_EQ(AllEntriesFor("foo"), "[ DEL, v1 ]");
1402    ASSERT_OK(dbfull()->TEST_CompactMemTable());  // Moves to level last-2
1403    ASSERT_EQ(AllEntriesFor("foo"), "[ DEL, v1 ]");
1404    dbfull()->TEST_CompactRange(last - 2, nullptr, nullptr);
1405    // DEL kept: "last" file overlaps
1406    ASSERT_EQ(AllEntriesFor("foo"), "[ DEL, v1 ]");
1407    dbfull()->TEST_CompactRange(last - 1, nullptr, nullptr);
1408    // Merging last-1 w/ last, so we are the base level for "foo", so
1409    // DEL is removed.  (as is v1).
1410    ASSERT_EQ(AllEntriesFor("foo"), "[ ]");
1411  }
1412  
1413  TEST(DBTest, OverlapInLevel0) {
1414    do {
1415      ASSERT_EQ(config::kMaxMemCompactLevel, 2) << "Fix test to match config";
1416  
1417      // Fill levels 1 and 2 to disable the pushing of new memtables to levels >
1418      // 0.
1419      ASSERT_OK(Put("100", "v100"));
1420      ASSERT_OK(Put("999", "v999"));
1421      dbfull()->TEST_CompactMemTable();
1422      ASSERT_OK(Delete("100"));
1423      ASSERT_OK(Delete("999"));
1424      dbfull()->TEST_CompactMemTable();
1425      ASSERT_EQ("0,1,1", FilesPerLevel());
1426  
1427      // Make files spanning the following ranges in level-0:
1428      //  files[0]  200 .. 900
1429      //  files[1]  300 .. 500
1430      // Note that files are sorted by smallest key.
1431      ASSERT_OK(Put("300", "v300"));
1432      ASSERT_OK(Put("500", "v500"));
1433      dbfull()->TEST_CompactMemTable();
1434      ASSERT_OK(Put("200", "v200"));
1435      ASSERT_OK(Put("600", "v600"));
1436      ASSERT_OK(Put("900", "v900"));
1437      dbfull()->TEST_CompactMemTable();
1438      ASSERT_EQ("2,1,1", FilesPerLevel());
1439  
1440      // Compact away the placeholder files we created initially
1441      dbfull()->TEST_CompactRange(1, nullptr, nullptr);
1442      dbfull()->TEST_CompactRange(2, nullptr, nullptr);
1443      ASSERT_EQ("2", FilesPerLevel());
1444  
1445      // Do a memtable compaction.  Before bug-fix, the compaction would
1446      // not detect the overlap with level-0 files and would incorrectly place
1447      // the deletion in a deeper level.
1448      ASSERT_OK(Delete("600"));
1449      dbfull()->TEST_CompactMemTable();
1450      ASSERT_EQ("3", FilesPerLevel());
1451      ASSERT_EQ("NOT_FOUND", Get("600"));
1452    } while (ChangeOptions());
1453  }
1454  
1455  TEST(DBTest, L0_CompactionBug_Issue44_a) {
1456    Reopen();
1457    ASSERT_OK(Put("b", "v"));
1458    Reopen();
1459    ASSERT_OK(Delete("b"));
1460    ASSERT_OK(Delete("a"));
1461    Reopen();
1462    ASSERT_OK(Delete("a"));
1463    Reopen();
1464    ASSERT_OK(Put("a", "v"));
1465    Reopen();
1466    Reopen();
1467    ASSERT_EQ("(a->v)", Contents());
1468    DelayMilliseconds(1000);  // Wait for compaction to finish
1469    ASSERT_EQ("(a->v)", Contents());
1470  }
1471  
1472  TEST(DBTest, L0_CompactionBug_Issue44_b) {
1473    Reopen();
1474    Put("", "");
1475    Reopen();
1476    Delete("e");
1477    Put("", "");
1478    Reopen();
1479    Put("c", "cv");
1480    Reopen();
1481    Put("", "");
1482    Reopen();
1483    Put("", "");
1484    DelayMilliseconds(1000);  // Wait for compaction to finish
1485    Reopen();
1486    Put("d", "dv");
1487    Reopen();
1488    Put("", "");
1489    Reopen();
1490    Delete("d");
1491    Delete("b");
1492    Reopen();
1493    ASSERT_EQ("(->)(c->cv)", Contents());
1494    DelayMilliseconds(1000);  // Wait for compaction to finish
1495    ASSERT_EQ("(->)(c->cv)", Contents());
1496  }
1497  
1498  TEST(DBTest, Fflush_Issue474) {
1499    static const int kNum = 100000;
1500    Random rnd(test::RandomSeed());
1501    for (int i = 0; i < kNum; i++) {
1502      fflush(nullptr);
1503      ASSERT_OK(Put(RandomKey(&rnd), RandomString(&rnd, 100)));
1504    }
1505  }
1506  
1507  TEST(DBTest, ComparatorCheck) {
1508    class NewComparator : public Comparator {
1509     public:
1510      const char* Name() const override { return "leveldb.NewComparator"; }
1511      int Compare(const Slice& a, const Slice& b) const override {
1512        return BytewiseComparator()->Compare(a, b);
1513      }
1514      void FindShortestSeparator(std::string* s, const Slice& l) const override {
1515        BytewiseComparator()->FindShortestSeparator(s, l);
1516      }
1517      void FindShortSuccessor(std::string* key) const override {
1518        BytewiseComparator()->FindShortSuccessor(key);
1519      }
1520    };
1521    NewComparator cmp;
1522    Options new_options = CurrentOptions();
1523    new_options.comparator = &cmp;
1524    Status s = TryReopen(&new_options);
1525    ASSERT_TRUE(!s.ok());
1526    ASSERT_TRUE(s.ToString().find("comparator") != std::string::npos)
1527        << s.ToString();
1528  }
1529  
1530  TEST(DBTest, CustomComparator) {
1531    class NumberComparator : public Comparator {
1532     public:
1533      const char* Name() const override { return "test.NumberComparator"; }
1534      int Compare(const Slice& a, const Slice& b) const override {
1535        return ToNumber(a) - ToNumber(b);
1536      }
1537      void FindShortestSeparator(std::string* s, const Slice& l) const override {
1538        ToNumber(*s);  // Check format
1539        ToNumber(l);   // Check format
1540      }
1541      void FindShortSuccessor(std::string* key) const override {
1542        ToNumber(*key);  // Check format
1543      }
1544  
1545     private:
1546      static int ToNumber(const Slice& x) {
1547        // Check that there are no extra characters.
1548        ASSERT_TRUE(x.size() >= 2 && x[0] == '[' && x[x.size() - 1] == ']')
1549            << EscapeString(x);
1550        int val;
1551        char ignored;
1552        ASSERT_TRUE(sscanf(x.ToString().c_str(), "[%i]%c", &val, &ignored) == 1)
1553            << EscapeString(x);
1554        return val;
1555      }
1556    };
1557    NumberComparator cmp;
1558    Options new_options = CurrentOptions();
1559    new_options.create_if_missing = true;
1560    new_options.comparator = &cmp;
1561    new_options.filter_policy = nullptr;   // Cannot use bloom filters
1562    new_options.write_buffer_size = 1000;  // Compact more often
1563    DestroyAndReopen(&new_options);
1564    ASSERT_OK(Put("[10]", "ten"));
1565    ASSERT_OK(Put("[0x14]", "twenty"));
1566    for (int i = 0; i < 2; i++) {
1567      ASSERT_EQ("ten", Get("[10]"));
1568      ASSERT_EQ("ten", Get("[0xa]"));
1569      ASSERT_EQ("twenty", Get("[20]"));
1570      ASSERT_EQ("twenty", Get("[0x14]"));
1571      ASSERT_EQ("NOT_FOUND", Get("[15]"));
1572      ASSERT_EQ("NOT_FOUND", Get("[0xf]"));
1573      Compact("[0]", "[9999]");
1574    }
1575  
1576    for (int run = 0; run < 2; run++) {
1577      for (int i = 0; i < 1000; i++) {
1578        char buf[100];
1579        snprintf(buf, sizeof(buf), "[%d]", i * 10);
1580        ASSERT_OK(Put(buf, buf));
1581      }
1582      Compact("[0]", "[1000000]");
1583    }
1584  }
1585  
1586  TEST(DBTest, ManualCompaction) {
1587    ASSERT_EQ(config::kMaxMemCompactLevel, 2)
1588        << "Need to update this test to match kMaxMemCompactLevel";
1589  
1590    MakeTables(3, "p", "q");
1591    ASSERT_EQ("1,1,1", FilesPerLevel());
1592  
1593    // Compaction range falls before files
1594    Compact("", "c");
1595    ASSERT_EQ("1,1,1", FilesPerLevel());
1596  
1597    // Compaction range falls after files
1598    Compact("r", "z");
1599    ASSERT_EQ("1,1,1", FilesPerLevel());
1600  
1601    // Compaction range overlaps files
1602    Compact("p1", "p9");
1603    ASSERT_EQ("0,0,1", FilesPerLevel());
1604  
1605    // Populate a different range
1606    MakeTables(3, "c", "e");
1607    ASSERT_EQ("1,1,2", FilesPerLevel());
1608  
1609    // Compact just the new range
1610    Compact("b", "f");
1611    ASSERT_EQ("0,0,2", FilesPerLevel());
1612  
1613    // Compact all
1614    MakeTables(1, "a", "z");
1615    ASSERT_EQ("0,1,2", FilesPerLevel());
1616    db_->CompactRange(nullptr, nullptr);
1617    ASSERT_EQ("0,0,1", FilesPerLevel());
1618  }
1619  
1620  TEST(DBTest, DBOpen_Options) {
1621    std::string dbname = test::TmpDir() + "/db_options_test";
1622    DestroyDB(dbname, Options());
1623  
1624    // Does not exist, and create_if_missing == false: error
1625    DB* db = nullptr;
1626    Options opts;
1627    opts.create_if_missing = false;
1628    Status s = DB::Open(opts, dbname, &db);
1629    ASSERT_TRUE(strstr(s.ToString().c_str(), "does not exist") != nullptr);
1630    ASSERT_TRUE(db == nullptr);
1631  
1632    // Does not exist, and create_if_missing == true: OK
1633    opts.create_if_missing = true;
1634    s = DB::Open(opts, dbname, &db);
1635    ASSERT_OK(s);
1636    ASSERT_TRUE(db != nullptr);
1637  
1638    delete db;
1639    db = nullptr;
1640  
1641    // Does exist, and error_if_exists == true: error
1642    opts.create_if_missing = false;
1643    opts.error_if_exists = true;
1644    s = DB::Open(opts, dbname, &db);
1645    ASSERT_TRUE(strstr(s.ToString().c_str(), "exists") != nullptr);
1646    ASSERT_TRUE(db == nullptr);
1647  
1648    // Does exist, and error_if_exists == false: OK
1649    opts.create_if_missing = true;
1650    opts.error_if_exists = false;
1651    s = DB::Open(opts, dbname, &db);
1652    ASSERT_OK(s);
1653    ASSERT_TRUE(db != nullptr);
1654  
1655    delete db;
1656    db = nullptr;
1657  }
1658  
1659  TEST(DBTest, DestroyEmptyDir) {
1660    std::string dbname = test::TmpDir() + "/db_empty_dir";
1661    TestEnv env(Env::Default());
1662    env.DeleteDir(dbname);
1663    ASSERT_TRUE(!env.FileExists(dbname));
1664  
1665    Options opts;
1666    opts.env = &env;
1667  
1668    ASSERT_OK(env.CreateDir(dbname));
1669    ASSERT_TRUE(env.FileExists(dbname));
1670    std::vector<std::string> children;
1671    ASSERT_OK(env.GetChildren(dbname, &children));
1672    // The stock Env's do not filter out '.' and '..' special files.
1673    ASSERT_EQ(2, children.size());
1674    ASSERT_OK(DestroyDB(dbname, opts));
1675    ASSERT_TRUE(!env.FileExists(dbname));
1676  
1677    // Should also be destroyed if Env is filtering out dot files.
1678    env.SetIgnoreDotFiles(true);
1679    ASSERT_OK(env.CreateDir(dbname));
1680    ASSERT_TRUE(env.FileExists(dbname));
1681    ASSERT_OK(env.GetChildren(dbname, &children));
1682    ASSERT_EQ(0, children.size());
1683    ASSERT_OK(DestroyDB(dbname, opts));
1684    ASSERT_TRUE(!env.FileExists(dbname));
1685  }
1686  
1687  TEST(DBTest, DestroyOpenDB) {
1688    std::string dbname = test::TmpDir() + "/open_db_dir";
1689    env_->DeleteDir(dbname);
1690    ASSERT_TRUE(!env_->FileExists(dbname));
1691  
1692    Options opts;
1693    opts.create_if_missing = true;
1694    DB* db = nullptr;
1695    ASSERT_OK(DB::Open(opts, dbname, &db));
1696    ASSERT_TRUE(db != nullptr);
1697  
1698    // Must fail to destroy an open db.
1699    ASSERT_TRUE(env_->FileExists(dbname));
1700    ASSERT_TRUE(!DestroyDB(dbname, Options()).ok());
1701    ASSERT_TRUE(env_->FileExists(dbname));
1702  
1703    delete db;
1704    db = nullptr;
1705  
1706    // Should succeed destroying a closed db.
1707    ASSERT_OK(DestroyDB(dbname, Options()));
1708    ASSERT_TRUE(!env_->FileExists(dbname));
1709  }
1710  
1711  TEST(DBTest, Locking) {
1712    DB* db2 = nullptr;
1713    Status s = DB::Open(CurrentOptions(), dbname_, &db2);
1714    ASSERT_TRUE(!s.ok()) << "Locking did not prevent re-opening db";
1715  }
1716  
1717  // Check that number of files does not grow when we are out of space
1718  TEST(DBTest, NoSpace) {
1719    Options options = CurrentOptions();
1720    options.env = env_;
1721    Reopen(&options);
1722  
1723    ASSERT_OK(Put("foo", "v1"));
1724    ASSERT_EQ("v1", Get("foo"));
1725    Compact("a", "z");
1726    const int num_files = CountFiles();
1727    // Force out-of-space errors.
1728    env_->no_space_.store(true, std::memory_order_release);
1729    for (int i = 0; i < 10; i++) {
1730      for (int level = 0; level < config::kNumLevels - 1; level++) {
1731        dbfull()->TEST_CompactRange(level, nullptr, nullptr);
1732      }
1733    }
1734    env_->no_space_.store(false, std::memory_order_release);
1735    ASSERT_LT(CountFiles(), num_files + 3);
1736  }
1737  
1738  TEST(DBTest, NonWritableFileSystem) {
1739    Options options = CurrentOptions();
1740    options.write_buffer_size = 1000;
1741    options.env = env_;
1742    Reopen(&options);
1743    ASSERT_OK(Put("foo", "v1"));
1744    // Force errors for new files.
1745    env_->non_writable_.store(true, std::memory_order_release);
1746    std::string big(100000, 'x');
1747    int errors = 0;
1748    for (int i = 0; i < 20; i++) {
1749      fprintf(stderr, "iter %d; errors %d\n", i, errors);
1750      if (!Put("foo", big).ok()) {
1751        errors++;
1752        DelayMilliseconds(100);
1753      }
1754    }
1755    ASSERT_GT(errors, 0);
1756    env_->non_writable_.store(false, std::memory_order_release);
1757  }
1758  
1759  TEST(DBTest, WriteSyncError) {
1760    // Check that log sync errors cause the DB to disallow future writes.
1761  
1762    // (a) Cause log sync calls to fail
1763    Options options = CurrentOptions();
1764    options.env = env_;
1765    Reopen(&options);
1766    env_->data_sync_error_.store(true, std::memory_order_release);
1767  
1768    // (b) Normal write should succeed
1769    WriteOptions w;
1770    ASSERT_OK(db_->Put(w, "k1", "v1"));
1771    ASSERT_EQ("v1", Get("k1"));
1772  
1773    // (c) Do a sync write; should fail
1774    w.sync = true;
1775    ASSERT_TRUE(!db_->Put(w, "k2", "v2").ok());
1776    ASSERT_EQ("v1", Get("k1"));
1777    ASSERT_EQ("NOT_FOUND", Get("k2"));
1778  
1779    // (d) make sync behave normally
1780    env_->data_sync_error_.store(false, std::memory_order_release);
1781  
1782    // (e) Do a non-sync write; should fail
1783    w.sync = false;
1784    ASSERT_TRUE(!db_->Put(w, "k3", "v3").ok());
1785    ASSERT_EQ("v1", Get("k1"));
1786    ASSERT_EQ("NOT_FOUND", Get("k2"));
1787    ASSERT_EQ("NOT_FOUND", Get("k3"));
1788  }
1789  
1790  TEST(DBTest, ManifestWriteError) {
1791    // Test for the following problem:
1792    // (a) Compaction produces file F
1793    // (b) Log record containing F is written to MANIFEST file, but Sync() fails
1794    // (c) GC deletes F
1795    // (d) After reopening DB, reads fail since deleted F is named in log record
1796  
1797    // We iterate twice.  In the second iteration, everything is the
1798    // same except the log record never makes it to the MANIFEST file.
1799    for (int iter = 0; iter < 2; iter++) {
1800      std::atomic<bool>* error_type = (iter == 0) ? &env_->manifest_sync_error_
1801                                                  : &env_->manifest_write_error_;
1802  
1803      // Insert foo=>bar mapping
1804      Options options = CurrentOptions();
1805      options.env = env_;
1806      options.create_if_missing = true;
1807      options.error_if_exists = false;
1808      DestroyAndReopen(&options);
1809      ASSERT_OK(Put("foo", "bar"));
1810      ASSERT_EQ("bar", Get("foo"));
1811  
1812      // Memtable compaction (will succeed)
1813      dbfull()->TEST_CompactMemTable();
1814      ASSERT_EQ("bar", Get("foo"));
1815      const int last = config::kMaxMemCompactLevel;
1816      ASSERT_EQ(NumTableFilesAtLevel(last), 1);  // foo=>bar is now in last level
1817  
1818      // Merging compaction (will fail)
1819      error_type->store(true, std::memory_order_release);
1820      dbfull()->TEST_CompactRange(last, nullptr, nullptr);  // Should fail
1821      ASSERT_EQ("bar", Get("foo"));
1822  
1823      // Recovery: should not lose data
1824      error_type->store(false, std::memory_order_release);
1825      Reopen(&options);
1826      ASSERT_EQ("bar", Get("foo"));
1827    }
1828  }
1829  
1830  TEST(DBTest, MissingSSTFile) {
1831    ASSERT_OK(Put("foo", "bar"));
1832    ASSERT_EQ("bar", Get("foo"));
1833  
1834    // Dump the memtable to disk.
1835    dbfull()->TEST_CompactMemTable();
1836    ASSERT_EQ("bar", Get("foo"));
1837  
1838    Close();
1839    ASSERT_TRUE(DeleteAnSSTFile());
1840    Options options = CurrentOptions();
1841    options.paranoid_checks = true;
1842    Status s = TryReopen(&options);
1843    ASSERT_TRUE(!s.ok());
1844    ASSERT_TRUE(s.ToString().find("issing") != std::string::npos) << s.ToString();
1845  }
1846  
1847  TEST(DBTest, StillReadSST) {
1848    ASSERT_OK(Put("foo", "bar"));
1849    ASSERT_EQ("bar", Get("foo"));
1850  
1851    // Dump the memtable to disk.
1852    dbfull()->TEST_CompactMemTable();
1853    ASSERT_EQ("bar", Get("foo"));
1854    Close();
1855    ASSERT_GT(RenameLDBToSST(), 0);
1856    Options options = CurrentOptions();
1857    options.paranoid_checks = true;
1858    Status s = TryReopen(&options);
1859    ASSERT_TRUE(s.ok());
1860    ASSERT_EQ("bar", Get("foo"));
1861  }
1862  
1863  TEST(DBTest, FilesDeletedAfterCompaction) {
1864    ASSERT_OK(Put("foo", "v2"));
1865    Compact("a", "z");
1866    const int num_files = CountFiles();
1867    for (int i = 0; i < 10; i++) {
1868      ASSERT_OK(Put("foo", "v2"));
1869      Compact("a", "z");
1870    }
1871    ASSERT_EQ(CountFiles(), num_files);
1872  }
1873  
1874  TEST(DBTest, BloomFilter) {
1875    env_->count_random_reads_ = true;
1876    Options options = CurrentOptions();
1877    options.env = env_;
1878    options.block_cache = NewLRUCache(0);  // Prevent cache hits
1879    options.filter_policy = NewBloomFilterPolicy(10);
1880    Reopen(&options);
1881  
1882    // Populate multiple layers
1883    const int N = 10000;
1884    for (int i = 0; i < N; i++) {
1885      ASSERT_OK(Put(Key(i), Key(i)));
1886    }
1887    Compact("a", "z");
1888    for (int i = 0; i < N; i += 100) {
1889      ASSERT_OK(Put(Key(i), Key(i)));
1890    }
1891    dbfull()->TEST_CompactMemTable();
1892  
1893    // Prevent auto compactions triggered by seeks
1894    env_->delay_data_sync_.store(true, std::memory_order_release);
1895  
1896    // Lookup present keys.  Should rarely read from small sstable.
1897    env_->random_read_counter_.Reset();
1898    for (int i = 0; i < N; i++) {
1899      ASSERT_EQ(Key(i), Get(Key(i)));
1900    }
1901    int reads = env_->random_read_counter_.Read();
1902    fprintf(stderr, "%d present => %d reads\n", N, reads);
1903    ASSERT_GE(reads, N);
1904    ASSERT_LE(reads, N + 2 * N / 100);
1905  
1906    // Lookup present keys.  Should rarely read from either sstable.
1907    env_->random_read_counter_.Reset();
1908    for (int i = 0; i < N; i++) {
1909      ASSERT_EQ("NOT_FOUND", Get(Key(i) + ".missing"));
1910    }
1911    reads = env_->random_read_counter_.Read();
1912    fprintf(stderr, "%d missing => %d reads\n", N, reads);
1913    ASSERT_LE(reads, 3 * N / 100);
1914  
1915    env_->delay_data_sync_.store(false, std::memory_order_release);
1916    Close();
1917    delete options.block_cache;
1918    delete options.filter_policy;
1919  }
1920  
1921  // Multi-threaded test:
1922  namespace {
1923  
1924  static const int kNumThreads = 4;
1925  static const int kTestSeconds = 10;
1926  static const int kNumKeys = 1000;
1927  
1928  struct MTState {
1929    DBTest* test;
1930    std::atomic<bool> stop;
1931    std::atomic<int> counter[kNumThreads];
1932    std::atomic<bool> thread_done[kNumThreads];
1933  };
1934  
1935  struct MTThread {
1936    MTState* state;
1937    int id;
1938  };
1939  
1940  static void MTThreadBody(void* arg) {
1941    MTThread* t = reinterpret_cast<MTThread*>(arg);
1942    int id = t->id;
1943    DB* db = t->state->test->db_;
1944    int counter = 0;
1945    fprintf(stderr, "... starting thread %d\n", id);
1946    Random rnd(1000 + id);
1947    std::string value;
1948    char valbuf[1500];
1949    while (!t->state->stop.load(std::memory_order_acquire)) {
1950      t->state->counter[id].store(counter, std::memory_order_release);
1951  
1952      int key = rnd.Uniform(kNumKeys);
1953      char keybuf[20];
1954      snprintf(keybuf, sizeof(keybuf), "%016d", key);
1955  
1956      if (rnd.OneIn(2)) {
1957        // Write values of the form <key, my id, counter>.
1958        // We add some padding for force compactions.
1959        snprintf(valbuf, sizeof(valbuf), "%d.%d.%-1000d", key, id,
1960                 static_cast<int>(counter));
1961        ASSERT_OK(db->Put(WriteOptions(), Slice(keybuf), Slice(valbuf)));
1962      } else {
1963        // Read a value and verify that it matches the pattern written above.
1964        Status s = db->Get(ReadOptions(), Slice(keybuf), &value);
1965        if (s.IsNotFound()) {
1966          // Key has not yet been written
1967        } else {
1968          // Check that the writer thread counter is >= the counter in the value
1969          ASSERT_OK(s);
1970          int k, w, c;
1971          ASSERT_EQ(3, sscanf(value.c_str(), "%d.%d.%d", &k, &w, &c)) << value;
1972          ASSERT_EQ(k, key);
1973          ASSERT_GE(w, 0);
1974          ASSERT_LT(w, kNumThreads);
1975          ASSERT_LE(c, t->state->counter[w].load(std::memory_order_acquire));
1976        }
1977      }
1978      counter++;
1979    }
1980    t->state->thread_done[id].store(true, std::memory_order_release);
1981    fprintf(stderr, "... stopping thread %d after %d ops\n", id, counter);
1982  }
1983  
1984  }  // namespace
1985  
1986  TEST(DBTest, MultiThreaded) {
1987    do {
1988      // Initialize state
1989      MTState mt;
1990      mt.test = this;
1991      mt.stop.store(false, std::memory_order_release);
1992      for (int id = 0; id < kNumThreads; id++) {
1993        mt.counter[id].store(false, std::memory_order_release);
1994        mt.thread_done[id].store(false, std::memory_order_release);
1995      }
1996  
1997      // Start threads
1998      MTThread thread[kNumThreads];
1999      for (int id = 0; id < kNumThreads; id++) {
2000        thread[id].state = &mt;
2001        thread[id].id = id;
2002        env_->StartThread(MTThreadBody, &thread[id]);
2003      }
2004  
2005      // Let them run for a while
2006      DelayMilliseconds(kTestSeconds * 1000);
2007  
2008      // Stop the threads and wait for them to finish
2009      mt.stop.store(true, std::memory_order_release);
2010      for (int id = 0; id < kNumThreads; id++) {
2011        while (!mt.thread_done[id].load(std::memory_order_acquire)) {
2012          DelayMilliseconds(100);
2013        }
2014      }
2015    } while (ChangeOptions());
2016  }
2017  
2018  namespace {
2019  typedef std::map<std::string, std::string> KVMap;
2020  }
2021  
2022  class ModelDB : public DB {
2023   public:
2024    class ModelSnapshot : public Snapshot {
2025     public:
2026      KVMap map_;
2027    };
2028  
2029    explicit ModelDB(const Options& options) : options_(options) {}
2030    ~ModelDB() override = default;
2031    Status Put(const WriteOptions& o, const Slice& k, const Slice& v) override {
2032      return DB::Put(o, k, v);
2033    }
2034    Status Delete(const WriteOptions& o, const Slice& key) override {
2035      return DB::Delete(o, key);
2036    }
2037    Status Get(const ReadOptions& options, const Slice& key,
2038               std::string* value) override {
2039      assert(false);  // Not implemented
2040      return Status::NotFound(key);
2041    }
2042    Iterator* NewIterator(const ReadOptions& options) override {
2043      if (options.snapshot == nullptr) {
2044        KVMap* saved = new KVMap;
2045        *saved = map_;
2046        return new ModelIter(saved, true);
2047      } else {
2048        const KVMap* snapshot_state =
2049            &(reinterpret_cast<const ModelSnapshot*>(options.snapshot)->map_);
2050        return new ModelIter(snapshot_state, false);
2051      }
2052    }
2053    const Snapshot* GetSnapshot() override {
2054      ModelSnapshot* snapshot = new ModelSnapshot;
2055      snapshot->map_ = map_;
2056      return snapshot;
2057    }
2058  
2059    void ReleaseSnapshot(const Snapshot* snapshot) override {
2060      delete reinterpret_cast<const ModelSnapshot*>(snapshot);
2061    }
2062    Status Write(const WriteOptions& options, WriteBatch* batch) override {
2063      class Handler : public WriteBatch::Handler {
2064       public:
2065        KVMap* map_;
2066        void Put(const Slice& key, const Slice& value) override {
2067          (*map_)[key.ToString()] = value.ToString();
2068        }
2069        void Delete(const Slice& key) override { map_->erase(key.ToString()); }
2070      };
2071      Handler handler;
2072      handler.map_ = &map_;
2073      return batch->Iterate(&handler);
2074    }
2075  
2076    bool GetProperty(const Slice& property, std::string* value) override {
2077      return false;
2078    }
2079    void GetApproximateSizes(const Range* r, int n, uint64_t* sizes) override {
2080      for (int i = 0; i < n; i++) {
2081        sizes[i] = 0;
2082      }
2083    }
2084    void CompactRange(const Slice* start, const Slice* end) override {}
2085  
2086   private:
2087    class ModelIter : public Iterator {
2088     public:
2089      ModelIter(const KVMap* map, bool owned)
2090          : map_(map), owned_(owned), iter_(map_->end()) {}
2091      ~ModelIter() override {
2092        if (owned_) delete map_;
2093      }
2094      bool Valid() const override { return iter_ != map_->end(); }
2095      void SeekToFirst() override { iter_ = map_->begin(); }
2096      void SeekToLast() override {
2097        if (map_->empty()) {
2098          iter_ = map_->end();
2099        } else {
2100          iter_ = map_->find(map_->rbegin()->first);
2101        }
2102      }
2103      void Seek(const Slice& k) override {
2104        iter_ = map_->lower_bound(k.ToString());
2105      }
2106      void Next() override { ++iter_; }
2107      void Prev() override { --iter_; }
2108      Slice key() const override { return iter_->first; }
2109      Slice value() const override { return iter_->second; }
2110      Status status() const override { return Status::OK(); }
2111  
2112     private:
2113      const KVMap* const map_;
2114      const bool owned_;  // Do we own map_
2115      KVMap::const_iterator iter_;
2116    };
2117    const Options options_;
2118    KVMap map_;
2119  };
2120  
2121  static bool CompareIterators(int step, DB* model, DB* db,
2122                               const Snapshot* model_snap,
2123                               const Snapshot* db_snap) {
2124    ReadOptions options;
2125    options.snapshot = model_snap;
2126    Iterator* miter = model->NewIterator(options);
2127    options.snapshot = db_snap;
2128    Iterator* dbiter = db->NewIterator(options);
2129    bool ok = true;
2130    int count = 0;
2131    for (miter->SeekToFirst(), dbiter->SeekToFirst();
2132         ok && miter->Valid() && dbiter->Valid(); miter->Next(), dbiter->Next()) {
2133      count++;
2134      if (miter->key().compare(dbiter->key()) != 0) {
2135        fprintf(stderr, "step %d: Key mismatch: '%s' vs. '%s'\n", step,
2136                EscapeString(miter->key()).c_str(),
2137                EscapeString(dbiter->key()).c_str());
2138        ok = false;
2139        break;
2140      }
2141  
2142      if (miter->value().compare(dbiter->value()) != 0) {
2143        fprintf(stderr, "step %d: Value mismatch for key '%s': '%s' vs. '%s'\n",
2144                step, EscapeString(miter->key()).c_str(),
2145                EscapeString(miter->value()).c_str(),
2146                EscapeString(miter->value()).c_str());
2147        ok = false;
2148      }
2149    }
2150  
2151    if (ok) {
2152      if (miter->Valid() != dbiter->Valid()) {
2153        fprintf(stderr, "step %d: Mismatch at end of iterators: %d vs. %d\n",
2154                step, miter->Valid(), dbiter->Valid());
2155        ok = false;
2156      }
2157    }
2158    fprintf(stderr, "%d entries compared: ok=%d\n", count, ok);
2159    delete miter;
2160    delete dbiter;
2161    return ok;
2162  }
2163  
2164  TEST(DBTest, Randomized) {
2165    Random rnd(test::RandomSeed());
2166    do {
2167      ModelDB model(CurrentOptions());
2168      const int N = 10000;
2169      const Snapshot* model_snap = nullptr;
2170      const Snapshot* db_snap = nullptr;
2171      std::string k, v;
2172      for (int step = 0; step < N; step++) {
2173        if (step % 100 == 0) {
2174          fprintf(stderr, "Step %d of %d\n", step, N);
2175        }
2176        // TODO(sanjay): Test Get() works
2177        int p = rnd.Uniform(100);
2178        if (p < 45) {  // Put
2179          k = RandomKey(&rnd);
2180          v = RandomString(
2181              &rnd, rnd.OneIn(20) ? 100 + rnd.Uniform(100) : rnd.Uniform(8));
2182          ASSERT_OK(model.Put(WriteOptions(), k, v));
2183          ASSERT_OK(db_->Put(WriteOptions(), k, v));
2184  
2185        } else if (p < 90) {  // Delete
2186          k = RandomKey(&rnd);
2187          ASSERT_OK(model.Delete(WriteOptions(), k));
2188          ASSERT_OK(db_->Delete(WriteOptions(), k));
2189  
2190        } else {  // Multi-element batch
2191          WriteBatch b;
2192          const int num = rnd.Uniform(8);
2193          for (int i = 0; i < num; i++) {
2194            if (i == 0 || !rnd.OneIn(10)) {
2195              k = RandomKey(&rnd);
2196            } else {
2197              // Periodically re-use the same key from the previous iter, so
2198              // we have multiple entries in the write batch for the same key
2199            }
2200            if (rnd.OneIn(2)) {
2201              v = RandomString(&rnd, rnd.Uniform(10));
2202              b.Put(k, v);
2203            } else {
2204              b.Delete(k);
2205            }
2206          }
2207          ASSERT_OK(model.Write(WriteOptions(), &b));
2208          ASSERT_OK(db_->Write(WriteOptions(), &b));
2209        }
2210  
2211        if ((step % 100) == 0) {
2212          ASSERT_TRUE(CompareIterators(step, &model, db_, nullptr, nullptr));
2213          ASSERT_TRUE(CompareIterators(step, &model, db_, model_snap, db_snap));
2214          // Save a snapshot from each DB this time that we'll use next
2215          // time we compare things, to make sure the current state is
2216          // preserved with the snapshot
2217          if (model_snap != nullptr) model.ReleaseSnapshot(model_snap);
2218          if (db_snap != nullptr) db_->ReleaseSnapshot(db_snap);
2219  
2220          Reopen();
2221          ASSERT_TRUE(CompareIterators(step, &model, db_, nullptr, nullptr));
2222  
2223          model_snap = model.GetSnapshot();
2224          db_snap = db_->GetSnapshot();
2225        }
2226      }
2227      if (model_snap != nullptr) model.ReleaseSnapshot(model_snap);
2228      if (db_snap != nullptr) db_->ReleaseSnapshot(db_snap);
2229    } while (ChangeOptions());
2230  }
2231  
2232  std::string MakeKey(unsigned int num) {
2233    char buf[30];
2234    snprintf(buf, sizeof(buf), "%016u", num);
2235    return std::string(buf);
2236  }
2237  
2238  void BM_LogAndApply(int iters, int num_base_files) {
2239    std::string dbname = test::TmpDir() + "/leveldb_test_benchmark";
2240    DestroyDB(dbname, Options());
2241  
2242    DB* db = nullptr;
2243    Options opts;
2244    opts.create_if_missing = true;
2245    Status s = DB::Open(opts, dbname, &db);
2246    ASSERT_OK(s);
2247    ASSERT_TRUE(db != nullptr);
2248  
2249    delete db;
2250    db = nullptr;
2251  
2252    Env* env = Env::Default();
2253  
2254    port::Mutex mu;
2255    MutexLock l(&mu);
2256  
2257    InternalKeyComparator cmp(BytewiseComparator());
2258    Options options;
2259    VersionSet vset(dbname, &options, nullptr, &cmp);
2260    bool save_manifest;
2261    ASSERT_OK(vset.Recover(&save_manifest));
2262    VersionEdit vbase;
2263    uint64_t fnum = 1;
2264    for (int i = 0; i < num_base_files; i++) {
2265      InternalKey start(MakeKey(2 * fnum), 1, kTypeValue);
2266      InternalKey limit(MakeKey(2 * fnum + 1), 1, kTypeDeletion);
2267      vbase.AddFile(2, fnum++, 1 /* file size */, start, limit);
2268    }
2269    ASSERT_OK(vset.LogAndApply(&vbase, &mu));
2270  
2271    uint64_t start_micros = env->NowMicros();
2272  
2273    for (int i = 0; i < iters; i++) {
2274      VersionEdit vedit;
2275      vedit.DeleteFile(2, fnum);
2276      InternalKey start(MakeKey(2 * fnum), 1, kTypeValue);
2277      InternalKey limit(MakeKey(2 * fnum + 1), 1, kTypeDeletion);
2278      vedit.AddFile(2, fnum++, 1 /* file size */, start, limit);
2279      vset.LogAndApply(&vedit, &mu);
2280    }
2281    uint64_t stop_micros = env->NowMicros();
2282    unsigned int us = stop_micros - start_micros;
2283    char buf[16];
2284    snprintf(buf, sizeof(buf), "%d", num_base_files);
2285    fprintf(stderr,
2286            "BM_LogAndApply/%-6s   %8d iters : %9u us (%7.0f us / iter)\n", buf,
2287            iters, us, ((float)us) / iters);
2288  }
2289  
2290  }  // namespace leveldb
2291  
2292  int main(int argc, char** argv) {
2293    if (argc > 1 && std::string(argv[1]) == "--benchmark") {
2294      leveldb::BM_LogAndApply(1000, 1);
2295      leveldb::BM_LogAndApply(1000, 100);
2296      leveldb::BM_LogAndApply(1000, 10000);
2297      leveldb::BM_LogAndApply(100, 100000);
2298      return 0;
2299    }
2300  
2301    return leveldb::test::RunAllTests();
2302  }