Support ingest_behind for IngestExternalFile

Summary:
First cut for early review; there are few conceptual points to answer and some code structure issues.

For conceptual points -

 - restriction-wise, we're going to disallow ingest_behind if (use_seqno_zero_out=true || disable_auto_compaction=false), the user is responsible to properly open and close DB with required params
 - we wanted to ingest into reserved bottom most level. Should we fail fast if bottom level isn't empty, or should we attempt to ingest if file fits there key-ranges-wise?
 - Modifying AssignLevelForIngestedFile seems the place we we'd handle that.

On code structure - going to refactor GenerateAndAddExternalFile call in the test class to allow passing instance of IngestionOptions, that's just going to incur lots of changes at callsites.
Closes https://github.com/facebook/rocksdb/pull/2144

Differential Revision: D4873732

Pulled By: lightmark

fbshipit-source-id: 81cb698106b68ef8797f564453651d50900e153a

db/column_family.cc

@@ -172,6 +172,12 @@ ColumnFamilyOptions SanitizeOptions(const ImmutableDBOptions& db_options,
       result.num_levels < 2) {
     result.num_levels = 2;
   }
+
+  if (result.compaction_style == kCompactionStyleUniversal &&
+      db_options.allow_ingest_behind && result.num_levels < 3) {
+    result.num_levels = 3;
+  }
+
   if (result.max_write_buffer_number < 2) {
     result.max_write_buffer_number = 2;
   }

db/compaction_iterator.cc

@@ -548,7 +548,8 @@ void CompactionIterator::PrepareOutput() {
 
   // This is safe for TransactionDB write-conflict checking since transactions
   // only care about sequence number larger than any active snapshots.
-  if (bottommost_level_ && valid_ && ikey_.sequence <= earliest_snapshot_ &&
+  if ((compaction_ != nullptr && !compaction_->allow_ingest_behind()) &&
+      bottommost_level_ && valid_ && ikey_.sequence <= earliest_snapshot_ &&
       ikey_.type != kTypeMerge &&
       !cmp_->Equal(compaction_->GetLargestUserKey(), ikey_.user_key)) {
     assert(ikey_.type != kTypeDeletion && ikey_.type != kTypeSingleDeletion);

db/compaction_iterator.h

@@ -18,6 +18,7 @@
 #include "db/merge_helper.h"
 #include "db/pinned_iterators_manager.h"
 #include "db/range_del_aggregator.h"
+#include "options/cf_options.h"
 #include "rocksdb/compaction_filter.h"
 
 namespace rocksdb {
@@ -48,6 +49,9 @@ class CompactionIterator {
     virtual Slice GetLargestUserKey() const {
       return compaction_->GetLargestUserKey();
     }
+    virtual bool allow_ingest_behind() const {
+      return compaction_->immutable_cf_options()->allow_ingest_behind;
+    }
 
    protected:
     CompactionProxy() = default;

db/compaction_iterator_test.cc

@@ -156,6 +156,7 @@ class FakeCompaction : public CompactionIterator::CompactionProxy {
   virtual Slice GetLargestUserKey() const {
     return "\xff\xff\xff\xff\xff\xff\xff\xff\xff";
   }
+  virtual bool allow_ingest_behind() const { return false; }
 
   bool key_not_exists_beyond_output_level = false;
 };

db/compaction_picker.cc

@@ -529,7 +529,11 @@ Compaction* CompactionPicker::CompactRange(
     // files together to the last level.
     assert(vstorage->num_levels() > 1);
     // DBImpl::CompactRange() set output level to be the last level
-    assert(output_level == vstorage->num_levels() - 1);
+    if (ioptions_.allow_ingest_behind) {
+      assert(output_level == vstorage->num_levels() - 2);
+    } else {
+      assert(output_level == vstorage->num_levels() - 1);
+    }
     // DBImpl::RunManualCompaction will make full range for universal compaction
     assert(begin == nullptr);
     assert(end == nullptr);

db/compaction_picker_test.cc

@@ -402,6 +402,35 @@ TEST_F(CompactionPickerTest, NeedsCompactionUniversal) {
               vstorage_->CompactionScore(0) >= 1);
   }
 }
+
+TEST_F(CompactionPickerTest, CompactionUniversalIngestBehindReservedLevel) {
+  const uint64_t kFileSize = 100000;
+  NewVersionStorage(1, kCompactionStyleUniversal);
+  ioptions_.allow_ingest_behind = true;
+  ioptions_.num_levels = 3;
+  UniversalCompactionPicker universal_compaction_picker(ioptions_, &icmp_);
+  // must return false when there's no files.
+  ASSERT_EQ(universal_compaction_picker.NeedsCompaction(vstorage_.get()),
+            false);
+
+  NewVersionStorage(3, kCompactionStyleUniversal);
+
+  Add(0, 1U, "150", "200", kFileSize, 0, 500, 550);
+  Add(0, 2U, "201", "250", kFileSize, 0, 401, 450);
+  Add(0, 4U, "260", "300", kFileSize, 0, 260, 300);
+  Add(1, 5U, "100", "151", kFileSize, 0, 200, 251);
+  Add(1, 3U, "301", "350", kFileSize, 0, 101, 150);
+  Add(2, 6U, "120", "200", kFileSize, 0, 20, 100);
+
+  UpdateVersionStorageInfo();
+
+  std::unique_ptr<Compaction> compaction(
+      universal_compaction_picker.PickCompaction(
+          cf_name_, mutable_cf_options_, vstorage_.get(), &log_buffer_));
+
+  // output level should be the one above the bottom-most
+  ASSERT_EQ(1, compaction->output_level());
+}
 // Tests if the files can be trivially moved in multi level
 // universal compaction when allow_trivial_move option is set
 // In this test as the input files overlaps, they cannot

db/compaction_picker_universal.cc

@@ -568,6 +568,13 @@ Compaction* UniversalCompactionPicker::PickCompactionToReduceSortedRuns(
     output_level = sorted_runs[first_index_after].level - 1;
   }
 
+  // last level is reserved for the files ingested behind
+  if (ioptions_.allow_ingest_behind &&
+      (output_level == vstorage->num_levels() - 1)) {
+    assert(output_level > 1);
+    output_level--;
+  }
+
   std::vector<CompactionInputFiles> inputs(vstorage->num_levels());
   for (size_t i = 0; i < inputs.size(); ++i) {
     inputs[i].level = start_level + static_cast<int>(i);
@@ -719,13 +726,20 @@ Compaction* UniversalCompactionPicker::PickCompactionToReduceSizeAmp(
                      cf_name.c_str(), file_num_buf);
   }
 
+  // output files at the bottom most level, unless it's reserved
+  int output_level = vstorage->num_levels() - 1;
+  // last level is reserved for the files ingested behind
+  if (ioptions_.allow_ingest_behind) {
+    assert(output_level > 1);
+    output_level--;
+  }
+
   return new Compaction(
       vstorage, ioptions_, mutable_cf_options, std::move(inputs),
-      vstorage->num_levels() - 1,
-      mutable_cf_options.MaxFileSizeForLevel(vstorage->num_levels() - 1),
+      output_level, mutable_cf_options.MaxFileSizeForLevel(output_level),
       /* max_grandparent_overlap_bytes */ LLONG_MAX, path_id,
       GetCompressionType(ioptions_, vstorage, mutable_cf_options,
-                         vstorage->num_levels() - 1, 1),
+                         output_level, 1),
       /* grandparents */ {}, /* is manual */ false, score,
       false /* deletion_compaction */,
       CompactionReason::kUniversalSizeAmplification);

db/db_impl.cc

@@ -2614,6 +2614,15 @@ Status DBImpl::IngestExternalFile(
   auto cfh = reinterpret_cast<ColumnFamilyHandleImpl*>(column_family);
   auto cfd = cfh->cfd();
 
+  // Ingest should immediately fail if ingest_behind is requested,
+  // but the DB doesn't support it.
+  if (ingestion_options.ingest_behind) {
+    if (!immutable_db_options_.allow_ingest_behind) {
+      return Status::InvalidArgument(
+        "Can't ingest_behind file in DB with allow_ingest_behind=false");
+    }
+  }
+
   ExternalSstFileIngestionJob ingestion_job(env_, versions_.get(), cfd,
                                             immutable_db_options_, env_options_,
                                             &snapshots_, ingestion_options);

db/db_impl_compaction_flush.cc

@@ -284,10 +284,14 @@ Status DBImpl::CompactRange(const CompactRangeOptions& options,
   if (cfd->ioptions()->compaction_style == kCompactionStyleUniversal &&
       cfd->NumberLevels() > 1) {
     // Always compact all files together.
+    final_output_level = cfd->NumberLevels() - 1;
+    // if bottom most level is reserved
+    if (immutable_db_options_.allow_ingest_behind) {
+      final_output_level--;
+    }
     s = RunManualCompaction(cfd, ColumnFamilyData::kCompactAllLevels,
-                            cfd->NumberLevels() - 1, options.target_path_id,
+                            final_output_level, options.target_path_id,
                             begin, end, exclusive);
-    final_output_level = cfd->NumberLevels() - 1;
   } else {
     for (int level = 0; level <= max_level_with_files; level++) {
       int output_level;

db/external_sst_file_ingestion_job.cc

@@ -171,9 +171,13 @@ Status ExternalSstFileIngestionJob::Run() {
 
   for (IngestedFileInfo& f : files_to_ingest_) {
     SequenceNumber assigned_seqno = 0;
-    status = AssignLevelAndSeqnoForIngestedFile(
-        super_version, force_global_seqno, cfd_->ioptions()->compaction_style,
-        &f, &assigned_seqno);
+    if (ingestion_options_.ingest_behind) {
+      status = CheckLevelForIngestedBehindFile(&f);
+    } else {
+      status = AssignLevelAndSeqnoForIngestedFile(
+         super_version, force_global_seqno, cfd_->ioptions()->compaction_style,
+         &f, &assigned_seqno);
+    }
     if (!status.ok()) {
       return status;
     }
@@ -408,9 +412,9 @@ Status ExternalSstFileIngestionJob::AssignLevelAndSeqnoForIngestedFile(
   Arena arena;
   ReadOptions ro;
   ro.total_order_seek = true;
-
   int target_level = 0;
   auto* vstorage = cfd_->current()->storage_info();
+
   for (int lvl = 0; lvl < cfd_->NumberLevels(); lvl++) {
     if (lvl > 0 && lvl < vstorage->base_level()) {
       continue;
@@ -418,20 +422,8 @@ Status ExternalSstFileIngestionJob::AssignLevelAndSeqnoForIngestedFile(
 
     if (vstorage->NumLevelFiles(lvl) > 0) {
       bool overlap_with_level = false;
-      MergeIteratorBuilder merge_iter_builder(&cfd_->internal_comparator(),
-                                              &arena);
-      RangeDelAggregator range_del_agg(cfd_->internal_comparator(),
-                                       {} /* snapshots */);
-      sv->current->AddIteratorsForLevel(ro, env_options_, &merge_iter_builder,
-                                        lvl, &range_del_agg);
-      if (!range_del_agg.IsEmpty()) {
-        return Status::NotSupported(
-            "file ingestion with range tombstones is currently unsupported");
-      }
-      ScopedArenaIterator level_iter(merge_iter_builder.Finish());
-
-      status = IngestedFileOverlapWithIteratorRange(
-          file_to_ingest, level_iter.get(), &overlap_with_level);
+      status = IngestedFileOverlapWithLevel(sv, file_to_ingest, lvl,
+        &overlap_with_level);
       if (!status.ok()) {
         return status;
       }
@@ -478,6 +470,33 @@ Status ExternalSstFileIngestionJob::AssignLevelAndSeqnoForIngestedFile(
   return status;
 }
 
+Status ExternalSstFileIngestionJob::CheckLevelForIngestedBehindFile(
+    IngestedFileInfo* file_to_ingest) {
+  auto* vstorage = cfd_->current()->storage_info();
+  // first check if new files fit in the bottommost level
+  int bottom_lvl = cfd_->NumberLevels() - 1;
+  if(!IngestedFileFitInLevel(file_to_ingest, bottom_lvl)) {
+    return Status::InvalidArgument(
+      "Can't ingest_behind file as it doesn't fit "
+      "at the bottommost level!");
+  }
+
+  // second check if despite allow_ingest_behind=true we still have 0 seqnums
+  // at some upper level
+  for (int lvl = 0; lvl < cfd_->NumberLevels() - 1; lvl++) {
+    for (auto file : vstorage->LevelFiles(lvl)) {
+      if (file->smallest_seqno == 0) {
+        return Status::InvalidArgument(
+          "Can't ingest_behind file as despite allow_ingest_behind=true "
+          "there are files with 0 seqno in database at upper levels!");
+      }
+    }
+  }
+
+  file_to_ingest->picked_level = bottom_lvl;
+  return Status::OK();
+}
+
 Status ExternalSstFileIngestionJob::AssignGlobalSeqnoForIngestedFile(
     IngestedFileInfo* file_to_ingest, SequenceNumber seqno) {
   if (file_to_ingest->original_seqno == seqno) {
@@ -564,6 +583,27 @@ bool ExternalSstFileIngestionJob::IngestedFileFitInLevel(
   return true;
 }
 
+Status ExternalSstFileIngestionJob::IngestedFileOverlapWithLevel(
+    SuperVersion* sv, IngestedFileInfo* file_to_ingest, int lvl,
+    bool* overlap_with_level) {
+  Arena arena;
+  ReadOptions ro;
+  ro.total_order_seek = true;
+  MergeIteratorBuilder merge_iter_builder(&cfd_->internal_comparator(),
+                                          &arena);
+  RangeDelAggregator range_del_agg(cfd_->internal_comparator(),
+                                   {} /* snapshots */);
+  sv->current->AddIteratorsForLevel(ro, env_options_, &merge_iter_builder,
+                                    lvl, &range_del_agg);
+  if (!range_del_agg.IsEmpty()) {
+    return Status::NotSupported(
+        "file ingestion with range tombstones is currently unsupported");
+  }
+  ScopedArenaIterator level_iter(merge_iter_builder.Finish());
+  return IngestedFileOverlapWithIteratorRange(
+      file_to_ingest, level_iter.get(), overlap_with_level);
+}
+
 }  // namespace rocksdb
 
 #endif  // !ROCKSDB_LITE

db/external_sst_file_ingestion_job.h

@@ -125,6 +125,12 @@ class ExternalSstFileIngestionJob {
                                             IngestedFileInfo* file_to_ingest,
                                             SequenceNumber* assigned_seqno);
 
+  // File that we want to ingest behind always goes to the lowest level;
+  // we just check that it fits in the level, that DB allows ingest_behind,
+  // and that we don't have 0 seqnums at the upper levels.
+  // REQUIRES: Mutex held
+  Status CheckLevelForIngestedBehindFile(IngestedFileInfo* file_to_ingest);
+
   // Set the file global sequence number to `seqno`
   Status AssignGlobalSeqnoForIngestedFile(IngestedFileInfo* file_to_ingest,
                                           SequenceNumber seqno);
@@ -135,6 +141,11 @@ class ExternalSstFileIngestionJob {
       const IngestedFileInfo* file_to_ingest, InternalIterator* iter,
       bool* overlap);
 
+  // Check if `file_to_ingest` key range overlap with level
+  // REQUIRES: Mutex held
+  Status IngestedFileOverlapWithLevel(SuperVersion* sv,
+    IngestedFileInfo* file_to_ingest, int lvl, bool* overlap_with_level);
+
   // Check if `file_to_ingest` can fit in level `level`
   // REQUIRES: Mutex held
   bool IngestedFileFitInLevel(const IngestedFileInfo* file_to_ingest,

db/external_sst_file_test.cc

@@ -90,6 +90,68 @@ class ExternalSSTFileTest : public DBTestBase {
     return s;
   }
 
+  Status GenerateAndAddExternalFileIngestBehind(
+      const Options options, const IngestExternalFileOptions ifo,
+      std::vector<std::pair<std::string, std::string>> data, int file_id = -1,
+      bool sort_data = false,
+      std::map<std::string, std::string>* true_data = nullptr,
+      ColumnFamilyHandle* cfh = nullptr) {
+    // Generate a file id if not provided
+    if (file_id == -1) {
+      file_id = last_file_id_ + 1;
+      last_file_id_++;
+    }
+
+    // Sort data if asked to do so
+    if (sort_data) {
+      std::sort(data.begin(), data.end(),
+                [&](const std::pair<std::string, std::string>& e1,
+                    const std::pair<std::string, std::string>& e2) {
+                  return options.comparator->Compare(e1.first, e2.first) < 0;
+                });
+      auto uniq_iter = std::unique(
+          data.begin(), data.end(),
+          [&](const std::pair<std::string, std::string>& e1,
+              const std::pair<std::string, std::string>& e2) {
+            return options.comparator->Compare(e1.first, e2.first) == 0;
+          });
+      data.resize(uniq_iter - data.begin());
+    }
+    std::string file_path = sst_files_dir_ + ToString(file_id);
+    SstFileWriter sst_file_writer(EnvOptions(), options, cfh);
+
+    Status s = sst_file_writer.Open(file_path);
+    if (!s.ok()) {
+      return s;
+    }
+    for (auto& entry : data) {
+      s = sst_file_writer.Add(entry.first, entry.second);
+      if (!s.ok()) {
+        sst_file_writer.Finish();
+        return s;
+      }
+    }
+    s = sst_file_writer.Finish();
+
+    if (s.ok()) {
+      if (cfh) {
+        s = db_->IngestExternalFile(cfh, {file_path}, ifo);
+      } else {
+        s = db_->IngestExternalFile({file_path}, ifo);
+      }
+    }
+
+    if (s.ok() && true_data) {
+      for (auto& entry : data) {
+        (*true_data)[entry.first] = entry.second;
+      }
+    }
+
+    return s;
+  }
+
+
+
   Status GenerateAndAddExternalFile(
       const Options options, std::vector<std::pair<int, std::string>> data,
       int file_id = -1, bool allow_global_seqno = false, bool sort_data = false,
@@ -1817,6 +1879,69 @@ TEST_F(ExternalSSTFileTest, SnapshotInconsistencyBug) {
 
   db_->ReleaseSnapshot(snap);
 }
+
+TEST_F(ExternalSSTFileTest, IngestBehind) {
+  Options options = CurrentOptions();
+  options.compaction_style = kCompactionStyleUniversal;
+  options.num_levels = 3;
+  options.disable_auto_compactions = false;
+  DestroyAndReopen(options);
+  std::vector<std::pair<std::string, std::string>> file_data;
+  std::map<std::string, std::string> true_data;
+
+  // Insert 100 -> 200 into the memtable
+  for (int i = 100; i <= 200; i++) {
+    ASSERT_OK(Put(Key(i), "memtable"));
+    true_data[Key(i)] = "memtable";
+  }
+
+  // Insert 100 -> 200 using IngestExternalFile
+  file_data.clear();
+  for (int i = 0; i <= 20; i++) {
+    file_data.emplace_back(Key(i), "ingest_behind");
+  }
+
+  IngestExternalFileOptions ifo;
+  ifo.allow_global_seqno = true;
+  ifo.ingest_behind = true;
+
+  // Can't ingest behind since allow_ingest_behind isn't set to true
+  ASSERT_NOK(GenerateAndAddExternalFileIngestBehind(options, ifo,
+                                                   file_data, -1, false,
+                                                   &true_data));
+
+  options.allow_ingest_behind = true;
+  // check that we still can open the DB, as num_levels should be
+  // sanitized to 3
+  options.num_levels = 2;
+  DestroyAndReopen(options);
+
+  options.num_levels = 3;
+  DestroyAndReopen(options);
+  // Insert 100 -> 200 into the memtable
+  for (int i = 100; i <= 200; i++) {
+    ASSERT_OK(Put(Key(i), "memtable"));
+    true_data[Key(i)] = "memtable";
+  }
+  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  // Universal picker should go at second from the bottom level
+  ASSERT_EQ("0,1", FilesPerLevel());
+  ASSERT_OK(GenerateAndAddExternalFileIngestBehind(options, ifo,
+                                                   file_data, -1, false,
+                                                   &true_data));
+  ASSERT_EQ("0,1,1", FilesPerLevel());
+  // this time ingest should fail as the file doesn't fit to the bottom level
+  ASSERT_NOK(GenerateAndAddExternalFileIngestBehind(options, ifo,
+                                                   file_data, -1, false,
+                                                   &true_data));
+  ASSERT_EQ("0,1,1", FilesPerLevel());
+  db_->CompactRange(CompactRangeOptions(), nullptr, nullptr);
+  // bottom level should be empty
+  ASSERT_EQ("0,1", FilesPerLevel());
+
+  size_t kcnt = 0;
+  VerifyDBFromMap(true_data, &kcnt, false);
+}
 }  // namespace rocksdb
 
 int main(int argc, char** argv) {

include/rocksdb/db.h

@@ -944,14 +944,22 @@ class DB {
   }
 
   // IngestExternalFile() will load a list of external SST files (1) into the DB
-  // We will try to find the lowest possible level that the file can fit in, and
-  // ingest the file into this level (2). A file that have a key range that
-  // overlap with the memtable key range will require us to Flush the memtable
-  // first before ingesting the file.
+  // Two primary modes are supported:
+  // - Duplicate keys in the new files will overwrite exiting keys (default)
+  // - Duplicate keys will be skipped (set ingest_behind=true)
+  // In the first mode we will try to find the lowest possible level that
+  // the file can fit in, and ingest the file into this level (2). A file that
+  // have a key range that overlap with the memtable key range will require us
+  // to Flush the memtable first before ingesting the file.
+  // In the second mode we will always ingest in the bottom mode level (see
+  // docs to IngestExternalFileOptions::ingest_behind).
   //
   // (1) External SST files can be created using SstFileWriter
   // (2) We will try to ingest the files to the lowest possible level
   //     even if the file compression dont match the level compression
+  // (3) If IngestExternalFileOptions->ingest_behind is set to true,
+  //     we always ingest at the bottommost level, which should be reserved
+  //     for this purpose (see DBOPtions::allow_ingest_behind flag).
   virtual Status IngestExternalFile(
       ColumnFamilyHandle* column_family,
       const std::vector<std::string>& external_files,

include/rocksdb/options.h

@@ -847,6 +847,18 @@ struct DBOptions {
   //
   // Dynamically changeable through SetDBOptions() API.
   bool avoid_flush_during_shutdown = false;
+
+  // Set this option to true during creation of database if you want
+  // to be able to ingest behind (call IngestExternalFile() skipping keys
+  // that already exist, rather than overwriting matching keys).
+  // Setting this option to true will affect 2 things:
+  // 1) Disable some internal optimizations around SST file compression
+  // 2) Reserve bottom-most level for ingested files only.
+  // 3) Note that num_levels should be >= 3 if this option is turned on.
+  //
+  // DEFAULT: false
+  // Immutable.
+  bool allow_ingest_behind = false;
 };
 
 // Options to control the behavior of a database (passed to DB::Open)
@@ -1126,6 +1138,14 @@ struct IngestExternalFileOptions {
   // If set to false and the file key range overlaps with the memtable key range
   // (memtable flush required), IngestExternalFile will fail.
   bool allow_blocking_flush = true;
+  // Set to true if you would like duplicate keys in the file being ingested
+  // to be skipped rather than overwriting existing data under that key.
+  // Usecase: back-fill of some historical data in the database without
+  // over-writing existing newer version of data.
+  // This option could only be used if the DB has been running
+  // with allow_ingest_behind=true since the dawn of time.
+  // All files will be ingested at the bottommost level with seqno=0.
+  bool ingest_behind = false;
 };
 
 }  // namespace rocksdb

options/cf_options.cc

@@ -71,6 +71,7 @@ ImmutableCFOptions::ImmutableCFOptions(const ImmutableDBOptions& db_options,
       num_levels(cf_options.num_levels),
       optimize_filters_for_hits(cf_options.optimize_filters_for_hits),
       force_consistency_checks(cf_options.force_consistency_checks),
+      allow_ingest_behind(db_options.allow_ingest_behind),
       listeners(db_options.listeners),
       row_cache(db_options.row_cache),
       max_subcompactions(db_options.max_subcompactions),

options/cf_options.h

@@ -108,6 +108,8 @@ struct ImmutableCFOptions {
 
   bool force_consistency_checks;
 
+  bool allow_ingest_behind;
+
   // A vector of EventListeners which call-back functions will be called
   // when specific RocksDB event happens.
   std::vector<std::shared_ptr<EventListener>> listeners;

options/db_options.cc

@@ -84,7 +84,8 @@ ImmutableDBOptions::ImmutableDBOptions(const DBOptions& options)
 #endif  // ROCKSDB_LITE
       fail_if_options_file_error(options.fail_if_options_file_error),
       dump_malloc_stats(options.dump_malloc_stats),
-      avoid_flush_during_recovery(options.avoid_flush_during_recovery) {
+      avoid_flush_during_recovery(options.avoid_flush_during_recovery),
+      allow_ingest_behind(options.allow_ingest_behind) {
 }
 
 void ImmutableDBOptions::Dump(Logger* log) const {
@@ -210,8 +211,11 @@ void ImmutableDBOptions::Dump(Logger* log) const {
   ROCKS_LOG_HEADER(log, "                             Options.wal_filter: %s",
                    wal_filter ? wal_filter->Name() : "None");
 #endif  // ROCKDB_LITE
+
   ROCKS_LOG_HEADER(log, "            Options.avoid_flush_during_recovery: %d",
                    avoid_flush_during_recovery);
+  ROCKS_LOG_HEADER(log, "            Options.allow_ingest_behind: %d",
+                   allow_ingest_behind);
 }
 
 MutableDBOptions::MutableDBOptions()

options/db_options.h

@@ -77,6 +77,7 @@ struct ImmutableDBOptions {
   bool fail_if_options_file_error;
   bool dump_malloc_stats;
   bool avoid_flush_during_recovery;
+  bool allow_ingest_behind;
 };
 
 struct MutableDBOptions {

options/options.cc

@@ -190,7 +190,8 @@ DBOptions::DBOptions(const Options& options)
       fail_if_options_file_error(options.fail_if_options_file_error),
       dump_malloc_stats(options.dump_malloc_stats),
       avoid_flush_during_recovery(options.avoid_flush_during_recovery),
-      avoid_flush_during_shutdown(options.avoid_flush_during_shutdown) {
+      avoid_flush_during_shutdown(options.avoid_flush_during_shutdown),
+      allow_ingest_behind(options.allow_ingest_behind) {
 }
 
 void DBOptions::Dump(Logger* log) const {

options/options_helper.cc

@@ -119,6 +119,8 @@ DBOptions BuildDBOptions(const ImmutableDBOptions& immutable_db_options,
       immutable_db_options.avoid_flush_during_recovery;
   options.avoid_flush_during_shutdown =
       mutable_db_options.avoid_flush_during_shutdown;
+  options.allow_ingest_behind =
+      immutable_db_options.allow_ingest_behind;
 
   return options;
 }

options/options_helper.h

@@ -335,7 +335,11 @@ static std::unordered_map<std::string, OptionTypeInfo> db_options_type_info = {
     {"avoid_flush_during_shutdown",
      {offsetof(struct DBOptions, avoid_flush_during_shutdown),
       OptionType::kBoolean, OptionVerificationType::kNormal, true,
-      offsetof(struct MutableDBOptions, avoid_flush_during_shutdown)}}};
+      offsetof(struct MutableDBOptions, avoid_flush_during_shutdown)}},
+    {"allow_ingest_behind",
+     {offsetof(struct DBOptions, allow_ingest_behind),
+      OptionType::kBoolean, OptionVerificationType::kNormal, false,
+      offsetof(struct ImmutableDBOptions, allow_ingest_behind)}}};
 
 // offset_of is used to get the offset of a class data member
 // ex: offset_of(&ColumnFamilyOptions::num_levels)

options/options_settable_test.cc

@@ -291,7 +291,8 @@ TEST_F(OptionsSettableTest, DBOptionsAllFieldsSettable) {
                              "dump_malloc_stats=false;"
                              "allow_2pc=false;"
                              "avoid_flush_during_recovery=false;"
-                             "avoid_flush_during_shutdown=false;",
+                             "avoid_flush_during_shutdown=false;"
+                             "allow_ingest_behind=false;",
                              new_options));
 
   ASSERT_EQ(unset_bytes_base, NumUnsetBytes(new_options_ptr, sizeof(DBOptions),