SERVER-48067 Reduce memory consumption for unique index builds with large numbers of non-unique keys

(cherry picked from commit 3c2951938d65667d675c48511d9d1046655809a5)

2 files changed, 16 insertions, 0 deletions

diff --git a/src/mongo/db/storage/key_string.cpp b/src/mongo/db/storage/key_string.cpp
index 627f804b2f4..b041cba850f 100644
--- a/src/mongo/db/storage/key_string.cpp
+++ b/src/mongo/db/storage/key_string.cpp
@@ -2564,6 +2564,15 @@ int compare(const char* leftBuf, const char* rightBuf, size_t leftSize, size_t r
     return leftSize < rightSize ? -1 : 1;
 }
 
+void Value::serializeWithoutRecordId(BufBuilder& buf) const {
+    dassert(decodeRecordIdAtEnd(_buffer.get(), _ksSize).isValid());
+
+    const int32_t sizeWithoutRecordId = sizeWithoutRecordIdAtEnd(_buffer.get(), _ksSize);
+    buf.appendNum(sizeWithoutRecordId);                          // Serialize size of KeyString
+    buf.appendBuf(_buffer.get(), sizeWithoutRecordId);           // Serialize KeyString
+    buf.appendBuf(_buffer.get() + _ksSize, _bufSize - _ksSize);  // Serialize TypeBits
+}
+
 template class BuilderBase<BufBuilder>;
 template class BuilderBase<StackBufBuilder>;
 
diff --git a/src/mongo/db/storage/key_string.h b/src/mongo/db/storage/key_string.h
index 33a57c2853a..b803c6ecf62 100644
--- a/src/mongo/db/storage/key_string.h
+++ b/src/mongo/db/storage/key_string.h
@@ -369,6 +369,13 @@ public:
         buf.appendBuf(_buffer.get(), _bufSize);  // Serialize Keystring + Typebits
     }
 
+    /**
+     * Serializes this Value, excluing the RecordId, into a storable format with TypeBits
+     * information. The serialized format takes the following form:
+     *   [keystring size][keystring encoding][typebits encoding]
+     */
+    void serializeWithoutRecordId(BufBuilder& buf) const;
+
     // Deserialize the Value from a serialized format.
     static Value deserialize(BufReader& buf, KeyString::Version version) {
         const int32_t sizeOfKeystring = buf.read<LittleEndian<int32_t>>();