@nxtedition/rocksdb 7.1.3 → 7.1.6

package/deps/rocksdb/rocksdb/db/compact_files_test.cc

@@ -77,7 +77,7 @@ TEST_F(CompactFilesTest, L0ConflictsFiles) {
   options.compression = kNoCompression;
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   assert(s.ok());
   assert(db);
@@ -128,7 +128,7 @@ TEST_F(CompactFilesTest, MultipleLevel) {
   options.listeners.emplace_back(collector);
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   ASSERT_OK(s);
   ASSERT_NE(db, nullptr);
@@ -211,7 +211,7 @@ TEST_F(CompactFilesTest, ObsoleteFiles) {
   options.listeners.emplace_back(collector);
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   ASSERT_OK(s);
   ASSERT_NE(db, nullptr);
@@ -250,7 +250,7 @@ TEST_F(CompactFilesTest, NotCutOutputOnLevel0) {
   options.listeners.emplace_back(collector);
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   assert(s.ok());
   assert(db);
@@ -288,7 +288,7 @@ TEST_F(CompactFilesTest, CapturingPendingFiles) {
   options.listeners.emplace_back(collector);
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   ASSERT_OK(s);
   assert(db);
@@ -366,7 +366,7 @@ TEST_F(CompactFilesTest, CompactionFilterWithGetSv) {
   options.compaction_filter = cf.get();
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   ASSERT_OK(s);
 
@@ -404,7 +404,7 @@ TEST_F(CompactFilesTest, SentinelCompressionType) {
        {CompactionStyle::kCompactionStyleLevel,
         CompactionStyle::kCompactionStyleUniversal,
         CompactionStyle::kCompactionStyleNone}) {
-    DestroyDB(db_name_, Options());
+    ASSERT_OK(DestroyDB(db_name_, Options()));
     Options options;
     options.compaction_style = compaction_style;
     // L0: Snappy, L1: ZSTD, L2: Snappy
@@ -458,7 +458,7 @@ TEST_F(CompactFilesTest, GetCompactionJobInfo) {
   options.listeners.emplace_back(collector);
 
   DB* db = nullptr;
-  DestroyDB(db_name_, options);
+  ASSERT_OK(DestroyDB(db_name_, options));
   Status s = DB::Open(options, db_name_, &db);
   ASSERT_OK(s);
   assert(db);

package/deps/rocksdb/rocksdb/db/compaction/compaction.cc

@@ -664,13 +664,6 @@ bool Compaction::ShouldFormSubcompactions() const {
     return false;
   }
 
-  // Note: the subcompaction boundary picking logic does not currently guarantee
-  // that all user keys that differ only by timestamp get processed by the same
-  // subcompaction.
-  if (cfd_->user_comparator()->timestamp_size() > 0) {
-    return false;
-  }
-
   // Round-Robin pri under leveled compaction allows subcompactions by default
   // and the number of subcompactions can be larger than max_subcompactions_
   if (cfd_->ioptions()->compaction_pri == kRoundRobin &&

package/deps/rocksdb/rocksdb/db/compaction/compaction.h

@@ -212,6 +212,11 @@ class Compaction {
   // Is this compaction creating a file in the bottom most level?
   bool bottommost_level() const { return bottommost_level_; }
 
+  // Is the compaction compact to the last level
+  bool is_last_level() const {
+    return output_level_ == immutable_options_.num_levels - 1;
+  }
+
   // Does this compaction include all sst files?
   bool is_full_compaction() const { return is_full_compaction_; }
 

package/deps/rocksdb/rocksdb/db/compaction/compaction_iterator.cc

@@ -80,6 +80,15 @@ CompactionIterator::CompactionIterator(
       compaction_filter_(compaction_filter),
       shutting_down_(shutting_down),
       manual_compaction_canceled_(manual_compaction_canceled),
+      bottommost_level_(!compaction_ ? false
+                                     : compaction_->bottommost_level() &&
+                                           !compaction_->allow_ingest_behind()),
+      // snapshots_ cannot be nullptr, but we will assert later in the body of
+      // the constructor.
+      visible_at_tip_(snapshots_ ? snapshots_->empty() : false),
+      earliest_snapshot_(!snapshots_ || snapshots_->empty()
+                             ? kMaxSequenceNumber
+                             : snapshots_->at(0)),
       info_log_(info_log),
       allow_data_in_errors_(allow_data_in_errors),
       enforce_single_del_contracts_(enforce_single_del_contracts),
@@ -98,25 +107,10 @@ CompactionIterator::CompactionIterator(
       level_(compaction_ == nullptr ? 0 : compaction_->level()),
       penultimate_level_cutoff_seqno_(penultimate_level_cutoff_seqno) {
   assert(snapshots_ != nullptr);
-  bottommost_level_ = compaction_ == nullptr
-                          ? false
-                          : compaction_->bottommost_level() &&
-                                !compaction_->allow_ingest_behind();
+
   if (compaction_ != nullptr) {
     level_ptrs_ = std::vector<size_t>(compaction_->number_levels(), 0);
   }
-  if (snapshots_->size() == 0) {
-    // optimize for fast path if there are no snapshots
-    visible_at_tip_ = true;
-    earliest_snapshot_iter_ = snapshots_->end();
-    earliest_snapshot_ = kMaxSequenceNumber;
-    latest_snapshot_ = 0;
-  } else {
-    visible_at_tip_ = false;
-    earliest_snapshot_iter_ = snapshots_->begin();
-    earliest_snapshot_ = snapshots_->at(0);
-    latest_snapshot_ = snapshots_->back();
-  }
 #ifndef NDEBUG
   // findEarliestVisibleSnapshot assumes this ordering.
   for (size_t i = 1; i < snapshots_->size(); ++i) {
@@ -173,7 +167,7 @@ void CompactionIterator::Next() {
       current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
       key_ = current_key_.GetInternalKey();
       ikey_.user_key = current_key_.GetUserKey();
-      valid_ = true;
+      validity_info_.SetValid(ValidContext::kMerge1);
     } else {
       // We consumed all pinned merge operands, release pinned iterators
       pinned_iters_mgr_.ReleasePinnedData();
@@ -191,7 +185,7 @@ void CompactionIterator::Next() {
     NextFromInput();
   }
 
-  if (valid_) {
+  if (Valid()) {
     // Record that we've outputted a record for the current key.
     has_outputted_key_ = true;
   }
@@ -228,7 +222,6 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
   {
     StopWatchNano timer(clock_, report_detailed_time_);
     if (kTypeBlobIndex == ikey_.type) {
-      blob_value_.Reset();
       filter = compaction_filter_->FilterBlobByKey(
           level_, filter_key, &compaction_filter_value_,
           compaction_filter_skip_until_.rep());
@@ -237,7 +230,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
         if (compaction_ == nullptr) {
           status_ =
               Status::Corruption("Unexpected blob index outside of compaction");
-          valid_ = false;
+          validity_info_.Invalidate();
           return false;
         }
 
@@ -252,7 +245,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
         Status s = blob_index.DecodeFrom(value_);
         if (!s.ok()) {
           status_ = s;
-          valid_ = false;
+          validity_info_.Invalidate();
           return false;
         }
 
@@ -270,7 +263,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
                                      &bytes_read);
         if (!s.ok()) {
           status_ = s;
-          valid_ = false;
+          validity_info_.Invalidate();
           return false;
         }
 
@@ -294,7 +287,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
     // Should not reach here, since FilterV2 should never return kUndetermined.
     status_ =
         Status::NotSupported("FilterV2() should never return kUndetermined");
-    valid_ = false;
+    validity_info_.Invalidate();
     return false;
   }
 
@@ -343,7 +336,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
       status_ = Status::NotSupported(
           "Only stacked BlobDB's internal compaction filter can return "
           "kChangeBlobIndex.");
-      valid_ = false;
+      validity_info_.Invalidate();
       return false;
     }
     if (ikey_.type == kTypeValue) {
@@ -356,7 +349,7 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
     if (!compaction_filter_->IsStackedBlobDbInternalCompactionFilter()) {
       status_ = Status::NotSupported(
           "CompactionFilter for integrated BlobDB should not return kIOError");
-      valid_ = false;
+      validity_info_.Invalidate();
       return false;
     }
     status_ = Status::IOError("Failed to access blob during compaction filter");
@@ -367,12 +360,13 @@ bool CompactionIterator::InvokeFilterIfNeeded(bool* need_skip,
 
 void CompactionIterator::NextFromInput() {
   at_next_ = false;
-  valid_ = false;
+  validity_info_.Invalidate();
 
-  while (!valid_ && input_.Valid() && !IsPausingManualCompaction() &&
+  while (!Valid() && input_.Valid() && !IsPausingManualCompaction() &&
          !IsShuttingDown()) {
     key_ = input_.key();
     value_ = input_.value();
+    blob_value_.Reset();
     iter_stats_.num_input_records++;
 
     Status pik_status = ParseInternalKey(key_, &ikey_, allow_data_in_errors_);
@@ -389,7 +383,7 @@ void CompactionIterator::NextFromInput() {
       has_current_user_key_ = false;
       current_user_key_sequence_ = kMaxSequenceNumber;
       current_user_key_snapshot_ = 0;
-      valid_ = true;
+      validity_info_.SetValid(ValidContext::kParseKeyError);
       break;
     }
     TEST_SYNC_POINT_CALLBACK("CompactionIterator:ProcessKV", &ikey_);
@@ -502,7 +496,7 @@ void CompactionIterator::NextFromInput() {
 
     if (UNLIKELY(!current_key_committed_)) {
       assert(snapshot_checker_ != nullptr);
-      valid_ = true;
+      validity_info_.SetValid(ValidContext::kCurrentKeyUncommitted);
       break;
     }
 
@@ -545,7 +539,7 @@ void CompactionIterator::NextFromInput() {
       }
 
       value_.clear();
-      valid_ = true;
+      validity_info_.SetValid(ValidContext::kKeepSDAndClearPut);
       clear_and_output_next_key_ = false;
     } else if (ikey_.type == kTypeSingleDeletion) {
       // We can compact out a SingleDelete if:
@@ -669,7 +663,7 @@ void CompactionIterator::NextFromInput() {
             ++iter_stats_.num_single_del_mismatch;
             if (enforce_single_del_contracts_) {
               ROCKS_LOG_ERROR(info_log_, "%s", oss.str().c_str());
-              valid_ = false;
+              validity_info_.Invalidate();
               status_ = Status::Corruption(oss.str());
               return;
             }
@@ -678,7 +672,7 @@ void CompactionIterator::NextFromInput() {
             // We cannot drop the SingleDelete as timestamp is enabled, and
             // timestamp of this key is greater than or equal to
             // *full_history_ts_low_. We will output the SingleDelete.
-            valid_ = true;
+            validity_info_.SetValid(ValidContext::kKeepTsHistory);
           } else if (has_outputted_key_ ||
                      DefinitelyInSnapshot(ikey_.sequence,
                                           earliest_write_conflict_snapshot_) ||
@@ -713,7 +707,7 @@ void CompactionIterator::NextFromInput() {
             // outputted on the next iteration.)
 
             // Setting valid_ to true will output the current SingleDelete
-            valid_ = true;
+            validity_info_.SetValid(ValidContext::kKeepSDForConflictCheck);
 
             // Set up the Put to be outputted in the next iteration.
             // (Optimization 3).
@@ -725,7 +719,7 @@ void CompactionIterator::NextFromInput() {
         } else {
           // We hit the next snapshot without hitting a put, so the iterator
           // returns the single delete.
-          valid_ = true;
+          validity_info_.SetValid(ValidContext::kKeepSDForSnapshot);
           TEST_SYNC_POINT_CALLBACK(
               "CompactionIterator::NextFromInput:SingleDelete:3",
               const_cast<Compaction*>(c));
@@ -758,11 +752,11 @@ void CompactionIterator::NextFromInput() {
           assert(bottommost_level_);
         } else {
           // Output SingleDelete
-          valid_ = true;
+          validity_info_.SetValid(ValidContext::kKeepSD);
         }
       }
 
-      if (valid_) {
+      if (Valid()) {
         at_next_ = true;
       }
     } else if (last_snapshot == current_user_key_snapshot_ ||
@@ -861,7 +855,7 @@ void CompactionIterator::NextFromInput() {
           (ParseInternalKey(input_.key(), &next_ikey, allow_data_in_errors_)
                .ok()) &&
           cmp_->EqualWithoutTimestamp(ikey_.user_key, next_ikey.user_key)) {
-        valid_ = true;
+        validity_info_.SetValid(ValidContext::kKeepDel);
         at_next_ = true;
       }
     } else if (ikey_.type == kTypeMerge) {
@@ -905,7 +899,7 @@ void CompactionIterator::NextFromInput() {
         current_key_.UpdateInternalKey(ikey_.sequence, ikey_.type);
         key_ = current_key_.GetInternalKey();
         ikey_.user_key = current_key_.GetUserKey();
-        valid_ = true;
+        validity_info_.SetValid(ValidContext::kMerge2);
       } else {
         // all merge operands were filtered out. reset the user key, since the
         // batch consumed by the merge operator should not shadow any keys
@@ -927,7 +921,7 @@ void CompactionIterator::NextFromInput() {
         ++iter_stats_.num_record_drop_range_del;
         AdvanceInputIter();
       } else {
-        valid_ = true;
+        validity_info_.SetValid(ValidContext::kNewUserKey);
       }
     }
 
@@ -936,13 +930,18 @@ void CompactionIterator::NextFromInput() {
     }
   }
 
-  if (!valid_ && IsShuttingDown()) {
+  if (!Valid() && IsShuttingDown()) {
     status_ = Status::ShutdownInProgress();
   }
 
   if (IsPausingManualCompaction()) {
     status_ = Status::Incomplete(Status::SubCode::kManualCompactionPaused);
   }
+
+  // Propagate corruption status from memtable itereator
+  if (!input_.Valid() && input_.status().IsCorruption()) {
+    status_ = input_.status();
+  }
 }
 
 bool CompactionIterator::ExtractLargeValueIfNeededImpl() {
@@ -955,7 +954,7 @@ bool CompactionIterator::ExtractLargeValueIfNeededImpl() {
 
   if (!s.ok()) {
     status_ = s;
-    valid_ = false;
+    validity_info_.Invalidate();
 
     return false;
   }
@@ -1000,7 +999,7 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
 
       if (!s.ok()) {
         status_ = s;
-        valid_ = false;
+        validity_info_.Invalidate();
 
         return;
       }
@@ -1026,7 +1025,7 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
 
       if (!s.ok()) {
         status_ = s;
-        valid_ = false;
+        validity_info_.Invalidate();
 
         return;
       }
@@ -1059,14 +1058,14 @@ void CompactionIterator::GarbageCollectBlobIfNeeded() {
     if (blob_decision == CompactionFilter::BlobDecision::kCorruption) {
       status_ =
           Status::Corruption("Corrupted blob reference encountered during GC");
-      valid_ = false;
+      validity_info_.Invalidate();
 
       return;
     }
 
     if (blob_decision == CompactionFilter::BlobDecision::kIOError) {
       status_ = Status::IOError("Could not relocate blob during GC");
-      valid_ = false;
+      validity_info_.Invalidate();
 
       return;
     }
@@ -1107,13 +1106,13 @@ void CompactionIterator::DecideOutputLevel() {
         compaction_->WithinPenultimateLevelOutputRange(ikey_.user_key);
     if (!safe_to_penultimate_level) {
       output_to_penultimate_level_ = false;
-      // It could happen when disable/enable `bottommost_temperature` while
-      // holding a snapshot. When `bottommost_temperature` is not set
+      // It could happen when disable/enable `last_level_temperature` while
+      // holding a snapshot. When `last_level_temperature` is not set
       // (==kUnknown), the data newer than any snapshot is pushed to the last
       // level, but when the per_key_placement feature is enabled on the fly,
       // the data later than the snapshot has to be moved to the penultimate
       // level, which may or may not be safe. So the user needs to make sure all
-      // snapshot is released before enabling `bottommost_temperature` feature
+      // snapshot is released before enabling `last_level_temperature` feature
       // We will migrate the feature to `last_level_temperature` and maybe make
       // it not dynamically changeable.
       if (ikey_.sequence > earliest_snapshot_) {
@@ -1126,7 +1125,7 @@ void CompactionIterator::DecideOutputLevel() {
 }
 
 void CompactionIterator::PrepareOutput() {
-  if (valid_) {
+  if (Valid()) {
     if (ikey_.type == kTypeValue) {
       ExtractLargeValueIfNeeded();
     } else if (ikey_.type == kTypeBlobIndex) {
@@ -1148,7 +1147,7 @@ void CompactionIterator::PrepareOutput() {
     //
     // Can we do the same for levels above bottom level as long as
     // KeyNotExistsBeyondOutputLevel() return true?
-    if (valid_ && compaction_ != nullptr &&
+    if (Valid() && compaction_ != nullptr &&
         !compaction_->allow_ingest_behind() && bottommost_level_ &&
         DefinitelyInSnapshot(ikey_.sequence, earliest_snapshot_) &&
         ikey_.type != kTypeMerge && current_key_committed_ &&
@@ -1162,13 +1161,14 @@ void CompactionIterator::PrepareOutput() {
             "earliest_snapshot %" PRIu64
             ", earliest_write_conflict_snapshot %" PRIu64
             " job_snapshot %" PRIu64
-            ". timestamp_size: %d full_history_ts_low_ %s",
+            ". timestamp_size: %d full_history_ts_low_ %s. validity %x",
             ikey_.DebugString(allow_data_in_errors_, true).c_str(),
             earliest_snapshot_, earliest_write_conflict_snapshot_,
             job_snapshot_, static_cast<int>(timestamp_size_),
             full_history_ts_low_ != nullptr
                 ? Slice(*full_history_ts_low_).ToString(true).c_str()
-                : "null");
+                : "null",
+            validity_info_.rep);
         assert(false);
       }
       ikey_.sequence = 0;
@@ -1284,7 +1284,10 @@ std::unique_ptr<BlobFetcher> CompactionIterator::CreateBlobFetcherIfNeeded(
     return nullptr;
   }
 
-  return std::unique_ptr<BlobFetcher>(new BlobFetcher(version, ReadOptions()));
+  ReadOptions read_options;
+  read_options.fill_cache = false;
+
+  return std::unique_ptr<BlobFetcher>(new BlobFetcher(version, read_options));
 }
 
 std::unique_ptr<PrefetchBufferCollection>

package/deps/rocksdb/rocksdb/db/compaction/compaction_iterator.h

@@ -235,7 +235,7 @@ class CompactionIterator {
   const Slice& value() const { return value_; }
   const Status& status() const { return status_; }
   const ParsedInternalKey& ikey() const { return ikey_; }
-  bool Valid() const { return valid_; }
+  inline bool Valid() const { return validity_info_.IsValid(); }
   const Slice& user_key() const { return current_user_key_; }
   const CompactionIterationStats& iter_stats() const { return iter_stats_; }
   uint64_t num_input_entry_scanned() const { return input_.num_itered(); }
@@ -332,29 +332,26 @@ class CompactionIterator {
   // earliest visible snapshot of an older value.
   // See WritePreparedTransactionTest::ReleaseSnapshotDuringCompaction3.
   std::unordered_set<SequenceNumber> released_snapshots_;
-  std::vector<SequenceNumber>::const_iterator earliest_snapshot_iter_;
   const SequenceNumber earliest_write_conflict_snapshot_;
   const SequenceNumber job_snapshot_;
   const SnapshotChecker* const snapshot_checker_;
   Env* env_;
   SystemClock* clock_;
-  bool report_detailed_time_;
-  bool expect_valid_internal_key_;
+  const bool report_detailed_time_;
+  const bool expect_valid_internal_key_;
   CompactionRangeDelAggregator* range_del_agg_;
   BlobFileBuilder* blob_file_builder_;
   std::unique_ptr<CompactionProxy> compaction_;
   const CompactionFilter* compaction_filter_;
   const std::atomic<bool>* shutting_down_;
   const std::atomic<bool>& manual_compaction_canceled_;
-  bool bottommost_level_;
-  bool valid_ = false;
-  bool visible_at_tip_;
-  SequenceNumber earliest_snapshot_;
-  SequenceNumber latest_snapshot_;
+  const bool bottommost_level_;
+  const bool visible_at_tip_;
+  const SequenceNumber earliest_snapshot_;
 
   std::shared_ptr<Logger> info_log_;
 
-  bool allow_data_in_errors_;
+  const bool allow_data_in_errors_;
 
   const bool enforce_single_del_contracts_;
 
@@ -370,8 +367,33 @@ class CompactionIterator {
 
   // State
   //
+  enum ValidContext : uint8_t {
+    kMerge1 = 0,
+    kMerge2 = 1,
+    kParseKeyError = 2,
+    kCurrentKeyUncommitted = 3,
+    kKeepSDAndClearPut = 4,
+    kKeepTsHistory = 5,
+    kKeepSDForConflictCheck = 6,
+    kKeepSDForSnapshot = 7,
+    kKeepSD = 8,
+    kKeepDel = 9,
+    kNewUserKey = 10,
+  };
+
+  struct ValidityInfo {
+    inline bool IsValid() const { return rep & 1; }
+    ValidContext GetContext() const {
+      return static_cast<ValidContext>(rep >> 1);
+    }
+    inline void SetValid(uint8_t ctx) { rep = (ctx << 1) | 1; }
+    inline void Invalidate() { rep = 0; }
+
+    uint8_t rep{0};
+  } validity_info_;
+
   // Points to a copy of the current compaction iterator output (current_key_)
-  // if valid_.
+  // if valid.
   Slice key_;
   // Points to the value in the underlying iterator that corresponds to the
   // current output.

package/deps/rocksdb/rocksdb/db/compaction/compaction_job.cc

@@ -240,8 +240,8 @@ void CompactionJob::Prepare() {
   bottommost_level_ = c->bottommost_level();
 
   if (c->ShouldFormSubcompactions()) {
-      StopWatch sw(db_options_.clock, stats_, SUBCOMPACTION_SETUP_TIME);
-      GenSubcompactionBoundaries();
+    StopWatch sw(db_options_.clock, stats_, SUBCOMPACTION_SETUP_TIME);
+    GenSubcompactionBoundaries();
   }
   if (boundaries_.size() > 1) {
     for (size_t i = 0; i <= boundaries_.size(); i++) {
@@ -250,6 +250,11 @@ void CompactionJob::Prepare() {
           (i != boundaries_.size()) ? std::optional<Slice>(boundaries_[i])
                                     : std::nullopt,
           static_cast<uint32_t>(i));
+      // assert to validate that boundaries don't have same user keys (without
+      // timestamp part).
+      assert(i == 0 || i == boundaries_.size() ||
+             cfd->user_comparator()->CompareWithoutTimestamp(
+                 boundaries_[i - 1], true, boundaries_[i], true) < 0);
     }
     RecordInHistogram(stats_, NUM_SUBCOMPACTIONS_SCHEDULED,
                       compact_->sub_compact_states.size());
@@ -479,9 +484,20 @@ void CompactionJob::GenSubcompactionBoundaries() {
   std::sort(
       all_anchors.begin(), all_anchors.end(),
       [cfd_comparator](TableReader::Anchor& a, TableReader::Anchor& b) -> bool {
-        return cfd_comparator->Compare(a.user_key, b.user_key) < 0;
+        return cfd_comparator->CompareWithoutTimestamp(a.user_key, true,
+                                                       b.user_key, true) < 0;
       });
 
+  // Remove duplicated entries from boundaries.
+  all_anchors.erase(
+      std::unique(all_anchors.begin(), all_anchors.end(),
+                  [cfd_comparator](TableReader::Anchor& a,
+                                   TableReader::Anchor& b) -> bool {
+                    return cfd_comparator->CompareWithoutTimestamp(
+                               a.user_key, true, b.user_key, true) == 0;
+                  }),
+      all_anchors.end());
+
   // Get the number of planned subcompactions, may update reserve threads
   // and update extra_num_subcompaction_threads_reserved_ for round-robin
   uint64_t num_planned_subcompactions;
@@ -508,6 +524,8 @@ void CompactionJob::GenSubcompactionBoundaries() {
       // of planned subcompactions
       num_planned_subcompactions =
           std::min(num_planned_subcompactions, GetSubcompactionsLimit());
+    } else {
+      num_planned_subcompactions = max_subcompactions_limit;
     }
   } else {
     num_planned_subcompactions = GetSubcompactionsLimit();
@@ -1022,6 +1040,9 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   const std::optional<Slice> start = sub_compact->start;
   const std::optional<Slice> end = sub_compact->end;
 
+  std::optional<Slice> start_without_ts;
+  std::optional<Slice> end_without_ts;
+
   ReadOptions read_options;
   read_options.verify_checksums = true;
   read_options.fill_cache = false;
@@ -1032,15 +1053,22 @@ void CompactionJob::ProcessKeyValueCompaction(SubcompactionState* sub_compact) {
   // (b) CompactionFilter::Decision::kRemoveAndSkipUntil.
   read_options.total_order_seek = true;
 
-  // Note: if we're going to support subcompactions for user-defined timestamps,
-  // the timestamp part will have to be stripped from the bounds here.
-  assert((!start.has_value() && !end.has_value()) ||
-         cfd->user_comparator()->timestamp_size() == 0);
+  // Remove the timestamps from boundaries because boundaries created in
+  // GenSubcompactionBoundaries doesn't strip away the timestamp.
+  size_t ts_sz = cfd->user_comparator()->timestamp_size();
   if (start.has_value()) {
     read_options.iterate_lower_bound = &start.value();
+    if (ts_sz > 0) {
+      start_without_ts = StripTimestampFromUserKey(start.value(), ts_sz);
+      read_options.iterate_lower_bound = &start_without_ts.value();
+    }
   }
   if (end.has_value()) {
     read_options.iterate_upper_bound = &end.value();
+    if (ts_sz > 0) {
+      end_without_ts = StripTimestampFromUserKey(end.value(), ts_sz);
+      read_options.iterate_upper_bound = &end_without_ts.value();
+    }
   }
 
   // Although the v2 aggregator is what the level iterator(s) know about,
@@ -1688,13 +1716,16 @@ Status CompactionJob::OpenCompactionOutputFile(SubcompactionState* sub_compact,
                            &syncpoint_arg);
 #endif
 
-  // Pass temperature of botommost files to FileSystem.
+  // Pass temperature of the last level files to FileSystem.
   FileOptions fo_copy = file_options_;
   Temperature temperature = sub_compact->compaction->output_temperature();
-  if (temperature == Temperature::kUnknown && bottommost_level_ &&
+  // only set for the last level compaction and also it's not output to
+  // penultimate level (when preclude_last_level feature is enabled)
+  if (temperature == Temperature::kUnknown &&
+      sub_compact->compaction->is_last_level() &&
       !sub_compact->IsCurrentPenultimateLevel()) {
     temperature =
-        sub_compact->compaction->mutable_cf_options()->bottommost_temperature;
+        sub_compact->compaction->mutable_cf_options()->last_level_temperature;
   }
   fo_copy.temperature = temperature;
 
@@ -1930,7 +1961,10 @@ void CompactionJob::LogCompaction() {
       stream.EndArray();
     }
     stream << "score" << compaction->score() << "input_data_size"
-           << compaction->CalculateTotalInputSize();
+           << compaction->CalculateTotalInputSize() << "oldest_snapshot_seqno"
+           << (existing_snapshots_.empty()
+                   ? int64_t{-1}  // Use -1 for "none"
+                   : static_cast<int64_t>(existing_snapshots_[0]));
   }
 }
 

package/deps/rocksdb/rocksdb/db/compaction/compaction_job_test.cc

@@ -372,8 +372,7 @@ class CompactionJobTestBase : public testing::Test {
                  smallest_seqno, largest_seqno, false, Temperature::kUnknown,
                  oldest_blob_file_number, kUnknownOldestAncesterTime,
                  kUnknownFileCreationTime, kUnknownFileChecksum,
-                 kUnknownFileChecksumFuncName, kDisableUserTimestamp,
-                 kDisableUserTimestamp, kNullUniqueId64x2);
+                 kUnknownFileChecksumFuncName, kNullUniqueId64x2);
 
     mutex_.Lock();
     EXPECT_OK(