update baseline benchmark

.idea/misc.xml

@@ -4,7 +4,6 @@
     <option name="/Default/RiderDebugger/RiderRestoreDecompile/RestoreDecompileSetting/@EntryValue" value="false" type="bool" />
     <option name="/Default/Housekeeping/GlobalSettingsUpgraded/IsUpgraded/@EntryValue" value="true" type="bool" />
     <option name="/Default/Housekeeping/FeatureSuggestion/FeatureSuggestionManager/DisabledSuggesters/=SwitchToGoToActionSuggester/@EntryIndexedValue" value="true" type="bool" />
-    <option name="/Default/Housekeeping/FeatureSuggestion/FeatureSuggestionManager/DisabledSuggesters/=SwitchToGoToActionSuggester/@EntryIndexRemoved" />
     <option name="/Default/Environment/Hierarchy/GeneratedFilesCacheKey/Timestamp/@EntryValue" value="5" type="long" />
   </component>
   <component name="CMakePythonSetting">

CMakeLists.txt

@@ -524,3 +524,8 @@ add_executable(test_fields
         "${PROJECT_SOURCE_DIR}/test/test_fields.cc"
 )
 target_link_libraries(test_fields PRIVATE leveldb gtest)
+
+add_executable(test_basicio
+        "${PROJECT_SOURCE_DIR}/test/test_basicio.cc"
+)
+target_link_libraries(test_basicio PRIVATE leveldb gtest)

benchmarks/db_bench.cc

@@ -74,7 +74,7 @@ static int FLAGS_reads = -1;
 static int FLAGS_threads = 1;
 
 // Size of each value
-static int FLAGS_value_size = 100;
+static int FLAGS_value_size = 1000;
 
 // Arrange to generate values that shrink to this fraction of
 // their original size after compression
@@ -852,8 +852,13 @@ class Benchmark {
       for (int j = 0; j < entries_per_batch_; j++) {
         const int k = seq ? i + j : thread->rand.Uniform(FLAGS_num);
         key.Set(k);
-        batch.Put(key.slice(), gen.Generate(value_size_));
-        bytes += value_size_ + key.slice().size();
+        auto value = gen.Generate(value_size_);
+        FieldArray field_array = {
+          {"1", value.ToString()},
+        };
+        auto encoded_value = Fields(field_array).Serialize();
+        batch.Put(key.slice(), Slice(encoded_value));
+        bytes += encoded_value.size() + key.slice().size();
         thread->stats.FinishedSingleOp();
       }
       s = db_->Write(write_options_, &batch);
@@ -870,7 +875,7 @@ class Benchmark {
     int i = 0;
     int64_t bytes = 0;
     for (iter->SeekToFirst(); i < reads_ && iter->Valid(); iter->Next()) {
-      bytes += iter->key().size() + iter->value().size();
+      bytes += iter->key().size() + iter->fields().size();
       thread->stats.FinishedSingleOp();
       ++i;
     }
@@ -883,7 +888,7 @@ class Benchmark {
     int i = 0;
     int64_t bytes = 0;
     for (iter->SeekToLast(); i < reads_ && iter->Valid(); iter->Prev()) {
-      bytes += iter->key().size() + iter->value().size();
+      bytes += iter->key().size() + iter->fields().size();
       thread->stats.FinishedSingleOp();
       ++i;
     }

db/db_impl.cc

@@ -257,6 +257,9 @@ void DBImpl::RemoveObsoleteFiles() {
         case kTableFile:
           keep = (live.find(number) != live.end());
           break;
+        case kVTableFile:
+          keep = (live.find(number) != live.end());
+          break;
         case kTempFile:
           // Any temp files that are currently being written to must
           // be recorded in pending_outputs_, which is inserted into "live"
@@ -1153,8 +1156,12 @@ Status DBImpl::Get(const ReadOptions& options, const Slice& key,
       s = current->Get(options, lkey, value, &stats);
       have_stat_update = true;
     }
-    auto fields = Fields(Slice(*value));
-    *value = fields["1"];
+    if (s.ok()) {
+      auto fields = Fields(Slice(*value));
+      *value = fields["1"];
+    } else {
+      *value = "";
+    }
     mutex_.Lock();
   }
 
@@ -1203,7 +1210,11 @@ Status DBImpl::Get(const ReadOptions& options, const Slice& key,
       s = current->Get(options, lkey, value, &stats);
       have_stat_update = true;
     }
-    *fields = Fields(Slice(*value));
+    if (s.ok()) {
+      *fields = Fields(Slice(*value));
+    } else {
+      *fields = Fields();
+    }
     mutex_.Lock();
   }
 

db/fields.cc

@@ -7,6 +7,7 @@ namespace leveldb {
      assert(!field_array.empty());
      for (const auto& field : field_array) {
        this->_fields[field.first] = field.second;
+       this->size_ += field.first.size() + field.second.size();
      }
    }
 
@@ -25,6 +26,7 @@ namespace leveldb {
       Slice value = Slice(field.data() + name_size, field.size() - name_size);
 
       this->_fields[name.ToString()] = value.ToString();
+      this->size_ += name.ToString().size() + value.ToString().size();
 
       fields = Slice(fields.data() + field_size, fields.size() - field_size);
     }

db/fields.h

@@ -33,8 +33,11 @@ namespace leveldb {
     // 将Fields编码为存入LevelDB的Value
     std::string Serialize() const;
 
+    uint64_t size() const { return size_; }
+
   private:
     std::map<std::string, std::string> _fields;
+    uint64_t size_ = 0;
   };
 }  // namespace leveldb
 #endif //STORAGE_LEVELDB_FIELDS_H_

db/filename.cc

@@ -112,6 +112,8 @@ bool ParseFileName(const std::string& filename, uint64_t* number,
       *type = kTableFile;
     } else if (suffix == Slice(".dbtmp")) {
       *type = kTempFile;
+    } else if (suffix == Slice(".vtb")) {
+      *type = kVTableFile;
     } else {
       return false;
     }

db/filename.h

@@ -25,7 +25,8 @@ enum FileType {
   kDescriptorFile,
   kCurrentFile,
   kTempFile,
-  kInfoLogFile  // Either the current one, or an old one
+  kInfoLogFile,  // Either the current one, or an old one
+  kVTableFile
 };
 
 // Return the name of the log file with the specified number

test/test_basicio.cc

@@ -0,0 +1,75 @@
+#include "gtest/gtest.h"
+#include "leveldb/env.h"
+#include "leveldb/db.h"
+using namespace leveldb;
+
+constexpr int value_size = 2048;
+constexpr int data_size = 128 << 20;
+
+std::map<std::string, std::string> data;
+
+Status OpenDB(std::string dbName, DB **db) {
+  Options options;
+  options.create_if_missing = true;
+  return DB::Open(options, dbName, db);
+}
+
+void InsertData(DB *db) {
+  WriteOptions writeOptions;
+  int key_num = data_size / value_size;
+  srand(0);
+
+  for (int i = 0; i < key_num; i++) {
+    int key_ = rand() % key_num+1;
+    std::string key = std::to_string(key_);
+    std::string value(value_size, 'a');
+    db->Put(writeOptions, key, value);
+    data[key] = value;
+  }
+}
+
+TEST(TestBasicIO, PointGet) {
+  DB *db;
+  if(OpenDB("testdb", &db).ok() == false) {
+    std::cerr << "open db failed" << std::endl;
+    abort();
+  }
+
+  data.clear();
+  InsertData(db);
+
+  ReadOptions readOptions;
+
+  for (auto iter = data.begin(); iter != data.end(); ++iter) {
+    std::string value;
+    db->Get(readOptions, iter->first, &value);
+    ASSERT_TRUE(value == iter->second);
+  }
+
+  delete db;
+}
+
+TEST(TestBasicIO, RangeQuery) {
+  DB *db;
+  if(OpenDB("testdb", &db).ok() == false) {
+    std::cerr << "open db failed" << std::endl;
+    abort();
+  }
+
+  data.clear();
+  InsertData(db);
+
+  ReadOptions readOptions;
+
+  auto iter = db->NewIterator(readOptions);
+  iter->SeekToFirst();
+  while (iter->Valid()) {
+    ASSERT_TRUE(data[iter->key().ToString()] == iter->fields()["1"]);
+    iter->Next();
+  }
+}
+
+int main(int argc, char **argv) {
+  testing::InitGoogleTest(&argc, argv);
+  return RUN_ALL_TESTS();
+}