From 355fa94365b4cf8b92938bee0becb18bc4526d82 Mon Sep 17 00:00:00 2001
From: SherlockNoMad <baihan.huang@gmail.com>
Date: Mon, 7 Dec 2015 10:51:08 -0800
Subject: [PATCH] EstimatedNumKeys Counter Inaccurate

---
 db/version_builder.cc |  3 ++-
 db/version_set.cc     | 33 +++++++++++++++++++++++++--------
 db/version_set.h      | 11 +++++++++--
 3 files changed, 36 insertions(+), 11 deletions(-)

diff --git a/db/version_builder.cc b/db/version_builder.cc
index 6717885b52..d6b1858a68 100644
--- a/db/version_builder.cc
+++ b/db/version_builder.cc
@@ -338,7 +338,8 @@ class VersionBuilder::Rep {
 
   void MaybeAddFile(VersionStorageInfo* vstorage, int level, FileMetaData* f) {
     if (levels_[level].deleted_files.count(f->fd.GetNumber()) > 0) {
-      // File is deleted: do nothing
+      // f is to-be-delected table file
+      vstorage->RemoveCurrentStats(f);
     } else {
       vstorage->AddFile(level, f, info_log_);
     }
diff --git a/db/version_set.cc b/db/version_set.cc
index 8d710f017d..ea8dde9eed 100644
--- a/db/version_set.cc
+++ b/db/version_set.cc
@@ -738,24 +738,24 @@ uint64_t VersionStorageInfo::GetEstimatedActiveKeys() const {
   // (2) keys are directly overwritten
   // (3) deletion on non-existing keys
   // (4) low number of samples
-  if (num_samples_ == 0) {
+  if (current_num_samples_ == 0) {
     return 0;
   }
 
-  if (accumulated_num_non_deletions_ <= accumulated_num_deletions_) {
+  if (current_num_non_deletions_ <= current_num_deletions_) {
     return 0;
   }
 
-  uint64_t est = accumulated_num_non_deletions_ - accumulated_num_deletions_;
+  uint64_t est = current_num_non_deletions_ - current_num_deletions_;
 
   uint64_t file_count = 0;
   for (int level = 0; level < num_levels_; ++level) {
     file_count += files_[level].size();
   }
 
-  if (num_samples_ < file_count) {
+  if (current_num_samples_ < file_count) {
     // casting to avoid overflowing
-    return (est * static_cast<double>(file_count) / num_samples_);
+    return (est * static_cast<double>(file_count) / current_num_samples_);
   } else {
     return est;
   }
@@ -826,7 +826,9 @@ VersionStorageInfo::VersionStorageInfo(
       accumulated_raw_value_size_(0),
       accumulated_num_non_deletions_(0),
       accumulated_num_deletions_(0),
-      num_samples_(0),
+      current_num_non_deletions_(0),
+      current_num_deletions_(0),
+      current_num_samples_(0),
       estimated_compaction_needed_bytes_(0),
       finalized_(false) {
   if (ref_vstorage != nullptr) {
@@ -836,7 +838,9 @@ VersionStorageInfo::VersionStorageInfo(
     accumulated_num_non_deletions_ =
         ref_vstorage->accumulated_num_non_deletions_;
     accumulated_num_deletions_ = ref_vstorage->accumulated_num_deletions_;
-    num_samples_ = ref_vstorage->num_samples_;
+    current_num_non_deletions_ = ref_vstorage->current_num_non_deletions_;
+    current_num_deletions_ = ref_vstorage->current_num_deletions_;
+    current_num_samples_ = ref_vstorage->current_num_samples_;
   }
 }
 
@@ -993,7 +997,20 @@ void VersionStorageInfo::UpdateAccumulatedStats(FileMetaData* file_meta) {
   accumulated_num_non_deletions_ +=
       file_meta->num_entries - file_meta->num_deletions;
   accumulated_num_deletions_ += file_meta->num_deletions;
-  num_samples_++;
+
+  current_num_non_deletions_ +=
+      file_meta->num_entries - file_meta->num_deletions;
+  current_num_deletions_ += file_meta->num_deletions;
+  current_num_samples_++;
+}
+
+void VersionStorageInfo::RemoveCurrentStats(FileMetaData* file_meta) {
+  if (file_meta->init_stats_from_file) {
+    current_num_non_deletions_ -=
+        file_meta->num_entries - file_meta->num_deletions;
+    current_num_deletions_ -= file_meta->num_deletions;
+    current_num_samples_--;
+  }
 }
 
 void Version::UpdateAccumulatedStats(bool update_stats) {
diff --git a/db/version_set.h b/db/version_set.h
index 0c0f3cc9dc..8d3d2c9c8e 100644
--- a/db/version_set.h
+++ b/db/version_set.h
@@ -111,6 +111,9 @@ class VersionStorageInfo {
   // Update the accumulated stats from a file-meta.
   void UpdateAccumulatedStats(FileMetaData* file_meta);
 
+  // Decrease the current stat form a to-be-delected file-meta
+  void RemoveCurrentStats(FileMetaData* file_meta);
+
   void ComputeCompensatedSizes();
 
   // Updates internal structures that keep track of compaction scores
@@ -400,8 +403,12 @@ class VersionStorageInfo {
   uint64_t accumulated_num_non_deletions_;
   // total number of deletion entries
   uint64_t accumulated_num_deletions_;
-  // the number of samples
-  uint64_t num_samples_;
+  // current number of non_deletion entries
+  uint64_t current_num_non_deletions_;
+  // current number of delection entries
+  uint64_t current_num_deletions_;
+  // current number of file samples
+  uint64_t current_num_samples_;
   // Estimated bytes needed to be compacted until all levels' size is down to
   // target sizes.
   uint64_t estimated_compaction_needed_bytes_;