liuyu/pgvector
1
0
Fork 0
mirror of https://github.com/pgvector/pgvector.git synced 2026-07-03 03:00:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

Fixed possible index corruption with HNSW vacuuming - resolves #988

Browse Source 
Co-authored-by: Bhagyesh Chaturvedi <bhagyeshc@google.com>
This commit is contained in:
Andrew Kane
2026-06-16 09:44:34 -07:00
parent 32284ba28a
commit eedba7ee14
3 changed files with 47 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
CHANGELOG.md
View File

@@ -1,5 +1,6 @@
## 0.8.3 (unreleased)
- Fixed possible index corruption with HNSW vacuuming
- Fixed performance regression with Hamming distance and Jaccard distance with Postgres 18
## 0.8.2 (2026-02-25)

1
src/hnsw.h
View File

@@ -412,6 +412,7 @@ typedef struct HnswVacuumState
BufferAccessStrategy bas;
HnswNeighborTuple ntup;
HnswElementData highestPoint;
HnswElementData fallbackPoint;
/* Memory */
MemoryContext tmpCtx;

55
src/hnswvacuum.c
View File

@@ -37,17 +37,20 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
{
BlockNumber blkno = HNSW_HEAD_BLKNO;
HnswElement highestPoint = &vacuumstate->highestPoint;
HnswElement fallbackPoint = &vacuumstate->fallbackPoint;
Relation index = vacuumstate->index;
BufferAccessStrategy bas = vacuumstate->bas;
HnswElement entryPoint = HnswGetEntryPoint(vacuumstate->index);
IndexBulkDeleteResult *stats = vacuumstate->stats;
/* Store separately since highestPoint.level is uint8 */
/* Store separately since HnswElement level is uint8 */
int highestLevel = -1;
int fallbackLevel = -1;
/* Initialize highest point */
/* Initialize highest point and fallback point */
highestPoint->blkno = InvalidBlockNumber;
highestPoint->offno = InvalidOffsetNumber;
fallbackPoint->blkno = InvalidBlockNumber;
fallbackPoint->offno = InvalidOffsetNumber;
while (BlockNumberIsValid(blkno))
{
@@ -119,14 +122,31 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
tidhash_insert(vacuumstate->deleted, ip, &found);
Assert(!found);
}
else if (etup->level > highestLevel && !(entryPoint != NULL && blkno == entryPoint->blkno && offno == entryPoint->offno))
else if (etup->level > highestLevel)
{
/* Keep track of highest non-entry point */
if (BlockNumberIsValid(highestPoint->blkno))
{
/* Current highest point becomes fallback */
fallbackPoint->blkno = highestPoint->blkno;
fallbackPoint->offno = highestPoint->offno;
fallbackPoint->level = highestPoint->level;
fallbackLevel = highestLevel;
}
/* Keep track of highest point */
highestPoint->blkno = blkno;
highestPoint->offno = offno;
highestPoint->level = etup->level;
highestLevel = etup->level;
}
else if (etup->level > fallbackLevel)
{
/* Keep track of second highest point */
fallbackPoint->blkno = blkno;
fallbackPoint->offno = offno;
fallbackPoint->level = etup->level;
fallbackLevel = etup->level;
}
}
blkno = HnswPageGetOpaque(page)->nextblkno;
@@ -269,12 +289,27 @@ RepairGraphEntryPoint(HnswVacuumState * vacuumstate)
/* Get a shared lock */
LockPage(index, HNSW_UPDATE_LOCK, ShareLock);
/* Load element */
HnswLoadElement(highestPoint, NULL, NULL, index, support, true, NULL);
/* Get latest entry point */
entryPoint = HnswGetEntryPoint(index);
/* Repair if needed */
if (NeedsUpdated(vacuumstate, highestPoint))
RepairGraphElement(vacuumstate, highestPoint, HnswGetEntryPoint(index));
/* Use fallback point if highest point is entry point */
if (entryPoint != NULL && entryPoint->blkno == highestPoint->blkno && entryPoint->offno == highestPoint->offno)
{
highestPoint = &vacuumstate->fallbackPoint;
if (!BlockNumberIsValid(highestPoint->blkno))
highestPoint = NULL;
}
if (highestPoint != NULL)
{
/* Load element */
HnswLoadElement(highestPoint, NULL, NULL, index, support, true, NULL);
/* Repair if needed */
if (NeedsUpdated(vacuumstate, highestPoint))
RepairGraphElement(vacuumstate, highestPoint, entryPoint);
}
/* Release lock */
UnlockPage(index, HNSW_UPDATE_LOCK, ShareLock);