Add EventListener::OnTableFileDeletion()

Summary:
Add EventListener::OnTableFileDeletion(), which will be
called when a table file is deleted.

Test Plan: Extend three existing tests in db_test to verify the deleted files.

Reviewers: rven, anthony, kradhakrishnan, igor, sdong

Reviewed By: sdong

Subscribers: dhruba, leveldb

Differential Revision: https://reviews.facebook.net/D38931

db/db_impl.cc

@@ -350,12 +350,15 @@ DBImpl::~DBImpl() {
   if (opened_successfully_) {
     JobContext job_context(next_job_id_.fetch_add(1));
     FindObsoleteFiles(&job_context, true);
+
+    mutex_.Unlock();
     // manifest number starting from 2
     job_context.manifest_file_number = 1;
     if (job_context.HaveSomethingToDelete()) {
       PurgeObsoleteFiles(job_context);
     }
     job_context.Clean();
+    mutex_.Lock();
   }
 
   for (auto l : logs_to_free_) {
@@ -520,7 +523,8 @@ void DBImpl::FindObsoleteFiles(JobContext* job_context, bool force,
     job_context->min_pending_output = std::numeric_limits<uint64_t>::max();
   }
 
-  // get obsolete files
+  // Get obsolete files.  This function will also update the list of
+  // pending files in VersionSet().
   versions_->GetObsoleteFiles(&job_context->sst_delete_files,
                               job_context->min_pending_output);
 
@@ -714,10 +718,10 @@ void DBImpl::PurgeObsoleteFiles(const JobContext& state) {
           file_deletion_status.ToString().c_str());
     }
     if (type == kTableFile) {
-      event_logger_.Log() << "job" << state.job_id << "event"
+      EventHelpers::LogAndNotifyTableFileDeletion(
-                          << "table_file_deletion"
+          &event_logger_, state.job_id, number, fname,
-                          << "file_number" << number
+          file_deletion_status, GetName(),
-                          << "status" << file_deletion_status.ToString();
+          db_options_.listeners);
     }
   }
 
@@ -751,10 +755,13 @@ void DBImpl::DeleteObsoleteFiles() {
   mutex_.AssertHeld();
   JobContext job_context(next_job_id_.fetch_add(1));
   FindObsoleteFiles(&job_context, true);
+
+  mutex_.Unlock();
   if (job_context.HaveSomethingToDelete()) {
     PurgeObsoleteFiles(job_context);
   }
   job_context.Clean();
+  mutex_.Lock();
 }
 
 Status DBImpl::Directories::CreateAndNewDirectory(
@@ -1433,7 +1440,7 @@ Status DBImpl::CompactFiles(
     // FindObsoleteFiles(). This is because job_context does not
     // catch all created files if compaction failed.
     FindObsoleteFiles(&job_context, !s.ok());
-  }
+  }  // release the mutex
 
   // delete unnecessary files if any, this is done outside the mutex
   if (job_context.HaveSomethingToDelete() || !log_buffer.IsEmpty()) {
@@ -1444,6 +1451,7 @@ Status DBImpl::CompactFiles(
     // It also applies to access other states that DB owns.
     log_buffer.FlushBufferToLog();
     if (job_context.HaveSomethingToDelete()) {
+      // no mutex is locked here.  No need to Unlock() and Lock() here.
       PurgeObsoleteFiles(job_context);
     }
     job_context.Clean();
@@ -3948,9 +3956,11 @@ Status DBImpl::DeleteFile(std::string name) {
     }
     FindObsoleteFiles(&job_context, false);
   }  // lock released here
+
   LogFlush(db_options_.info_log);
   // remove files outside the db-lock
   if (job_context.HaveSomethingToDelete()) {
+    // Call PurgeObsoleteFiles() without holding mutex.
     PurgeObsoleteFiles(job_context);
   }
   job_context.Clean();

db/db_test.cc

@@ -11400,6 +11400,38 @@ TEST_F(DBTest, MigrateToDynamicLevelMaxBytesBase) {
   ASSERT_EQ(NumTableFilesAtLevel(2), 0);
 }
 
+namespace {
+class OnFileDeletionListener : public EventListener {
+ public:
+  OnFileDeletionListener() :
+      matched_count_(0),
+      expected_file_name_("") {}
+
+  void SetExpectedFileName(
+      const std::string file_name) {
+    expected_file_name_ = file_name;
+  }
+
+  void VerifyMatchedCount(size_t expected_value) {
+    ASSERT_EQ(matched_count_, expected_value);
+  }
+
+  void OnTableFileDeleted(
+      const TableFileDeletionInfo& info) override {
+    if (expected_file_name_ != "") {
+      ASSERT_EQ(expected_file_name_, info.file_path);
+      expected_file_name_ = "";
+      matched_count_++;
+    }
+  }
+
+ private:
+  size_t matched_count_;
+  std::string expected_file_name_;
+};
+
+}  // namespace
+
 TEST_F(DBTest, DynamicLevelCompressionPerLevel) {
   if (!Snappy_Supported()) {
     return;
@@ -11432,6 +11464,9 @@ TEST_F(DBTest, DynamicLevelCompressionPerLevel) {
   options.compression_per_level[1] = kNoCompression;
   options.compression_per_level[2] = kSnappyCompression;
 
+  OnFileDeletionListener* listener = new OnFileDeletionListener();
+  options.listeners.emplace_back(listener);
+
   DestroyAndReopen(options);
 
   // Insert more than 80K. L4 should be base level. Neither L0 nor L4 should
@@ -11464,8 +11499,11 @@ TEST_F(DBTest, DynamicLevelCompressionPerLevel) {
   ColumnFamilyMetaData cf_meta;
   db_->GetColumnFamilyMetaData(&cf_meta);
   for (auto file : cf_meta.levels[4].files) {
+    listener->SetExpectedFileName(dbname_ + file.name);
     ASSERT_OK(dbfull()->DeleteFile(file.name));
   }
+  listener->VerifyMatchedCount(cf_meta.levels[4].files.size());
+
   int num_keys = 0;
   std::unique_ptr<Iterator> iter(db_->NewIterator(ReadOptions()));
   for (iter->SeekToFirst(); iter->Valid(); iter->Next()) {
@@ -12162,6 +12200,8 @@ TEST_F(DBTest, DeleteMovedFileAfterCompaction) {
     options.create_if_missing = true;
     options.level0_file_num_compaction_trigger =
         2;  // trigger compaction when we have 2 files
+    OnFileDeletionListener* listener = new OnFileDeletionListener();
+    options.listeners.emplace_back(listener);
     DestroyAndReopen(options);
 
     Random rnd(301);
@@ -12214,12 +12254,14 @@ TEST_F(DBTest, DeleteMovedFileAfterCompaction) {
     ASSERT_EQ("0,0,2", FilesPerLevel(0));
 
     // iterator is holding the file
-    ASSERT_TRUE(env_->FileExists(dbname_ + "/" + moved_file_name));
+    ASSERT_TRUE(env_->FileExists(dbname_ + moved_file_name));
 
+    listener->SetExpectedFileName(dbname_ + moved_file_name);
     iterator.reset();
 
     // this file should have been compacted away
-    ASSERT_TRUE(!env_->FileExists(dbname_ + "/" + moved_file_name));
+    ASSERT_TRUE(!env_->FileExists(dbname_ + moved_file_name));
+    listener->VerifyMatchedCount(1);
   }
 }
 
@@ -12393,6 +12435,10 @@ TEST_F(DBTest, DeleteObsoleteFilesPendingOutputs) {
       2;  // trigger compaction when we have 2 files
   options.max_background_flushes = 2;
   options.max_background_compactions = 2;
+
+  OnFileDeletionListener* listener = new OnFileDeletionListener();
+  options.listeners.emplace_back(listener);
+
   Reopen(options);
 
   Random rnd(301);
@@ -12441,6 +12487,7 @@ TEST_F(DBTest, DeleteObsoleteFilesPendingOutputs) {
   db_->GetLiveFilesMetaData(&metadata);
   ASSERT_EQ(metadata.size(), 1U);
   auto file_on_L2 = metadata[0].name;
+  listener->SetExpectedFileName(dbname_ + file_on_L2);
 
   ASSERT_OK(dbfull()->TEST_CompactRange(3, nullptr, nullptr));
   ASSERT_EQ("0,0,0,0,1", FilesPerLevel(0));
@@ -12456,7 +12503,8 @@ TEST_F(DBTest, DeleteObsoleteFilesPendingOutputs) {
   ASSERT_EQ(metadata.size(), 2U);
 
   // This file should have been deleted
-  ASSERT_TRUE(!env_->FileExists(dbname_ + "/" + file_on_L2));
+  ASSERT_TRUE(!env_->FileExists(dbname_ + file_on_L2));
+  listener->VerifyMatchedCount(1);
 }
 
 TEST_F(DBTest, CloseSpeedup) {

db/event_helpers.cc

@@ -73,4 +73,36 @@ void EventHelpers::LogAndNotifyTableFileCreation(
 #endif  // !ROCKSDB_LITE
 }
 
+void EventHelpers::LogAndNotifyTableFileDeletion(
+    EventLogger* event_logger, int job_id,
+    uint64_t file_number, const std::string& file_path,
+    const Status& status, const std::string& dbname,
+    const std::vector<std::shared_ptr<EventListener>>& listeners) {
+
+  JSONWriter jwriter;
+  AppendCurrentTime(&jwriter);
+
+  jwriter << "job" << job_id
+          << "event" << "table_file_deletion"
+          << "file_number" << file_number;
+  if (!status.ok()) {
+    jwriter << "status" << status.ToString();
+  }
+
+  jwriter.EndObject();
+
+  event_logger->Log(jwriter);
+
+#ifndef ROCKSDB_LITE
+  TableFileDeletionInfo info;
+  info.db_name = dbname;
+  info.job_id = job_id;
+  info.file_path = file_path;
+  info.status = status;
+  for (auto listener : listeners) {
+    listener->OnTableFileDeleted(info);
+  }
+#endif  // !ROCKSDB_LITE
+}
+
 }  // namespace rocksdb

db/event_helpers.h

@@ -23,6 +23,11 @@ class EventHelpers {
       EventLogger* event_logger,
       const std::vector<std::shared_ptr<EventListener>>& listeners,
       const FileDescriptor& fd, const TableFileCreationInfo& info);
+  static void LogAndNotifyTableFileDeletion(
+      EventLogger* event_logger, int job_id,
+      uint64_t file_number, const std::string& file_path,
+      const Status& status, const std::string& db_name,
+      const std::vector<std::shared_ptr<EventListener>>& listeners);
 };
 
 }  // namespace rocksdb

include/rocksdb/listener.h

@@ -39,6 +39,17 @@ struct TableFileCreationInfo {
 
 #ifndef ROCKSDB_LITE
 
+struct TableFileDeletionInfo {
+  // The name of the database where the file was deleted.
+  std::string db_name;
+  // The path to the deleted file.
+  std::string file_path;
+  // The id of the job which deleted the file.
+  int job_id;
+  // The status indicating whether the deletion was successfull or not.
+  Status status;
+};
+
 struct CompactionJobInfo {
   CompactionJobInfo() = default;
   explicit CompactionJobInfo(const CompactionJobStats& _stats) :
@@ -122,6 +133,20 @@ class EventListener {
       bool triggered_writes_slowdown,
       bool triggered_writes_stop) {}
 
+  // A call-back function for RocksDB which will be called whenever
+  // a SST file is deleted.  Different from OnCompactionCompleted and
+  // OnFlushCompleted, this call-back is designed for external logging
+  // service and thus only provide string parameters instead
+  // of a pointer to DB.  Applications that build logic basic based
+  // on file creations and deletions is suggested to implement
+  // OnFlushCompleted and OnCompactionCompleted.
+  //
+  // Note that if applications would like to use the passed reference
+  // outside this function call, they should make copies from the
+  // returned value.
+  virtual void OnTableFileDeleted(
+      const TableFileDeletionInfo& info) {}
+
   // A call-back function for RocksDB which will be called whenever
   // a registered RocksDB compacts a file. The default implementation
   // is a no-op.