[feature-wip](unique-key-merge-on-write) update bitmap after compacti…

…on, DSIP-018 (#11289)
apache · Jul 30, 2022 · 2783267 · 2783267
1 parent 9333e79
commit 2783267
Show file tree

Hide file tree

Showing 4 changed files with 53 additions and 0 deletions.
diff --git a/be/src/olap/compaction.cpp b/be/src/olap/compaction.cpp
@@ -252,6 +252,14 @@ Status Compaction::modify_rowsets() {
     std::vector<RowsetSharedPtr> output_rowsets;
     output_rowsets.push_back(_output_rowset);
     std::lock_guard<std::shared_mutex> wrlock(_tablet->get_header_lock());
+
+    // update dst rowset delete bitmap
+    if (_tablet->keys_type() == KeysType::UNIQUE_KEYS &&
+        _tablet->enable_unique_key_merge_on_write()) {
+        _tablet->tablet_meta()->update_delete_bitmap(_input_rowsets, _output_rs_writer->version(),
+                                                     _rowid_conversion);
+    }
+
     RETURN_NOT_OK(_tablet->modify_rowsets(output_rowsets, _input_rowsets, true));
     _tablet->save_meta();
     return Status::OK();

diff --git a/be/src/olap/merger.cpp b/be/src/olap/merger.cpp
@@ -103,6 +103,7 @@ Status Merger::vmerge_rowsets(TabletSharedPtr tablet, ReaderType reader_type,
     reader_params.rs_readers = src_rowset_readers;
     reader_params.version = dst_rowset_writer->version();
     reader_params.tablet_schema = cur_tablet_schema;
+    reader_params.delete_bitmap = &tablet->tablet_meta()->delete_bitmap();
     if (stats_output && stats_output->rowid_conversion) {
         reader_params.record_rowids = true;
     }

diff --git a/be/src/olap/tablet_meta.cpp b/be/src/olap/tablet_meta.cpp
@@ -670,6 +670,8 @@ void TabletMeta::delete_stale_rs_meta_by_version(const Version& version) {
     while (it != _stale_rs_metas.end()) {
         if ((*it)->version() == version) {
             it = _stale_rs_metas.erase(it);
+            // remove rowset delete bitmap
+            delete_bitmap().remove({(*it)->rowset_id(), 0, 0}, {(*it)->rowset_id(), UINT32_MAX, 0});
         } else {
             it++;
         }
@@ -755,6 +757,44 @@ Status TabletMeta::set_partition_id(int64_t partition_id) {
     return Status::OK();
 }
 
+// We take a delete bitmap's snapshot of origin rowset at the beginning of
+// compaction, some keys of origin rowsets might be deleted during compaction,
+// but exist in dest rowset. so we need to update the bitmap of dest rowset
+// after compaction.
+// ANNT: should take a tablet lock before calling the function
+void TabletMeta::update_delete_bitmap(const std::vector<RowsetSharedPtr>& input_rowsets,
+                                      const Version& version,
+                                      const RowIdConversion& rowid_conversion) {
+    RowLocation src;
+    RowLocation dst;
+    DeleteBitmap output_rowset_delete_bitmap(_tablet_id);
+    for (auto& rowset : input_rowsets) {
+        src.rowset_id = rowset->rowset_id();
+        for (uint32_t seg_id = 0; seg_id < rowset->num_segments(); ++seg_id) {
+            src.segment_id = seg_id;
+            DeleteBitmap upper_map(_tablet_id);
+            delete_bitmap().subset({rowset->rowset_id(), seg_id, version.second},
+                                   {rowset->rowset_id(), seg_id, INT64_MAX}, &upper_map);
+            // traverse all versions and convert rowid
+            for (auto iter = upper_map.delete_bitmap.begin(); iter != upper_map.delete_bitmap.end();
+                 ++iter) {
+                auto cur_version = std::get<2>(iter->first);
+                for (auto index = iter->second.begin(); index != iter->second.end(); ++index) {
+                    src.row_id = *index;
+                    if (rowid_conversion.get(src, &dst) != 0) {
+                        LOG(WARNING) << "Can't find rowid, may be deleted by the delete_handler.";
+                        continue;
+                    }
+                    output_rowset_delete_bitmap.add({dst.rowset_id, dst.segment_id, cur_version},
+                                                    dst.row_id);
+                }
+            }
+        }
+    }
+    // update output rowset delete bitmap
+    delete_bitmap().merge(output_rowset_delete_bitmap);
+}
+
 bool operator==(const TabletMeta& a, const TabletMeta& b) {
     if (a._table_id != b._table_id) return false;
     if (a._partition_id != b._partition_id) return false;

diff --git a/be/src/olap/tablet_meta.h b/be/src/olap/tablet_meta.h
@@ -29,6 +29,7 @@
 #include "olap/delete_handler.h"
 #include "olap/olap_common.h"
 #include "olap/olap_define.h"
+#include "olap/rowid_conversion.h"
 #include "olap/rowset/rowset.h"
 #include "olap/rowset/rowset_meta.h"
 #include "olap/tablet_schema.h"
@@ -205,6 +206,9 @@ class TabletMeta {
 
     bool enable_unique_key_merge_on_write() const { return _enable_unique_key_merge_on_write; }
 
+    void update_delete_bitmap(const std::vector<RowsetSharedPtr>& input_rowsets,
+                              const Version& version, const RowIdConversion& rowid_conversion);
+
 private:
     Status _save_meta(DataDir* data_dir);