Version bump to 0.8.0 [skip ci]

This reverts commit 7043cce893.

.github/workflows/build.yml

@@ -8,8 +8,8 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          # - postgres: 18
-          #   os: ubuntu-24.04
+          - postgres: 18
+            os: ubuntu-24.04
           - postgres: 17
             os: ubuntu-24.04
           - postgres: 16

CHANGELOG.md

@@ -1,10 +1,10 @@
-## 0.8.0 (unreleased)
+## 0.8.0 (2024-10-30)
 
 - Added support for iterative index scans
 - Added casts for arrays to `sparsevec`
-- Improved cost estimation
+- Improved cost estimation for better index selection when filtering
+- Improved performance of HNSW index scans
 - Improved performance of HNSW inserts and on-disk index builds
-- Reduced memory usage for HNSW index scans
 - Dropped support for Postgres 12
 
 ## 0.7.4 (2024-08-05)

META.json

@@ -2,7 +2,7 @@
 	"name": "vector",
 	"abstract": "Open-source vector similarity search for Postgres",
 	"description": "Supports L2 distance, inner product, and cosine distance",
-	"version": "0.7.4",
+	"version": "0.8.0",
 	"maintainer": [
 		"Andrew Kane <andrew@ankane.org>"
 	],
@@ -12,7 +12,7 @@
 	"prereqs": {
 		"runtime": {
 			"requires": {
-				"PostgreSQL": "12.0.0"
+				"PostgreSQL": "13.0.0"
 			}
 		}
 	},
@@ -20,7 +20,7 @@
 		"vector": {
 			"file": "sql/vector.sql",
 			"docfile": "README.md",
-			"version": "0.7.4",
+			"version": "0.8.0",
 			"abstract": "Open-source vector similarity search for Postgres"
 		}
 	},

Makefile

@@ -1,5 +1,5 @@
 EXTENSION = vector
-EXTVERSION = 0.7.4
+EXTVERSION = 0.8.0
 
 MODULE_big = vector
 DATA = $(wildcard sql/*--*--*.sql)

Makefile.win

@@ -1,5 +1,5 @@
 EXTENSION = vector
-EXTVERSION = 0.7.4
+EXTVERSION = 0.8.0
 
 DATA_built = sql\$(EXTENSION)--$(EXTVERSION).sql
 OBJS = src\bitutils.obj src\bitvec.obj src\halfutils.obj src\halfvec.obj src\hnsw.obj src\hnswbuild.obj src\hnswinsert.obj src\hnswscan.obj src\hnswutils.obj src\hnswvacuum.obj src\ivfbuild.obj src\ivfflat.obj src\ivfinsert.obj src\ivfkmeans.obj src\ivfscan.obj src\ivfutils.obj src\ivfvacuum.obj src\sparsevec.obj src\vector.obj

README.md

@@ -17,11 +17,11 @@ Plus [ACID](https://en.wikipedia.org/wiki/ACID) compliance, point-in-time recove
 
 ### Linux and Mac
 
-Compile and install the extension (supports Postgres 12+)
+Compile and install the extension (supports Postgres 13+)
 
 ```sh
 cd /tmp
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 make
 make install # may need sudo
@@ -46,7 +46,7 @@ Then use `nmake` to build:
 ```cmd
 set "PGROOT=C:\Program Files\PostgreSQL\16"
 cd %TEMP%
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 nmake /F Makefile.win
 nmake /F Makefile.win install
@@ -324,7 +324,7 @@ For a large number of workers, you may also need to increase `max_parallel_worke
 
 ### Indexing Progress
 
-Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING)
 
 ```sql
 SELECT phase, round(100.0 * blocks_done / nullif(blocks_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -410,7 +410,7 @@ For a large number of workers, you may also need to increase `max_parallel_worke
 
 ### Indexing Progress
 
-Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
+Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING)
 
 ```sql
 SELECT phase, round(100.0 * tuples_done / nullif(tuples_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
@@ -427,30 +427,126 @@ Note: `%` is only populated during the `loading tuples` phase
 
 ## Filtering
 
-There are a few ways to index nearest neighbor queries with a `WHERE` clause
+There are a few ways to index nearest neighbor queries with a `WHERE` clause.
 
 ```sql
 SELECT * FROM items WHERE category_id = 123 ORDER BY embedding <-> '[3,1,2]' LIMIT 5;
 ```
 
-Create an index on one [or more](https://www.postgresql.org/docs/current/indexes-multicolumn.html) of the `WHERE` columns for exact search
+A good place to start is creating an index on the filter column. This can provide fast, exact nearest neighbor search in many cases. Postgres has a number of [index types](https://www.postgresql.org/docs/current/indexes-types.html) for this: B-tree (default), hash, GiST, SP-GiST, GIN, and BRIN.
 
 ```sql
 CREATE INDEX ON items (category_id);
 ```
 
-Or a [partial index](https://www.postgresql.org/docs/current/indexes-partial.html) on the vector column for approximate search
+For multiple columns, consider a [multicolumn index](https://www.postgresql.org/docs/current/indexes-multicolumn.html).
+
+```sql
+CREATE INDEX ON items (location_id, category_id);
+```
+
+Exact indexes work well for conditions that match a low percentage of rows. Otherwise, [approximate indexes](#indexing) can work better.
+
+```sql
+CREATE INDEX ON items USING hnsw (embedding vector_l2_ops);
+```
+
+With approximate indexes, filtering is applied *after* the index is scanned. If a condition matches 10% of rows, with HNSW and the default `hnsw.ef_search` of 40, only 4 rows will match on average. For more rows, increase `hnsw.ef_search`.
+
+```sql
+SET hnsw.ef_search = 200;
+```
+
+Starting with 0.8.0, you can enable [iterative index scans](#iterative-index-scans), which will automatically scan more of the index when needed.
+
+```sql
+SET hnsw.iterative_scan = strict_order;
+```
+
+If filtering by only a few distinct values, consider [partial indexing](https://www.postgresql.org/docs/current/indexes-partial.html).
 
 ```sql
 CREATE INDEX ON items USING hnsw (embedding vector_l2_ops) WHERE (category_id = 123);
 ```
 
-Use [partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html) for approximate search on many different values of the `WHERE` columns
+If filtering by many different values, consider [partitioning](https://www.postgresql.org/docs/current/ddl-partitioning.html).
 
 ```sql
 CREATE TABLE items (embedding vector(3), category_id int) PARTITION BY LIST(category_id);
 ```
 
+## Iterative Index Scans
+
+*Added in 0.8.0*
+
+With approximate indexes, queries with filtering can return less results since filtering is applied *after* the index is scanned. Starting with 0.8.0, you can enable iterative index scans, which will automatically scan more of the index until enough results are found (or it reaches `hnsw.max_scan_tuples` or `ivfflat.max_probes`).
+
+Iterative scans can use strict or relaxed ordering.
+
+Strict ensures results are in the exact order by distance
+
+```sql
+SET hnsw.iterative_scan = strict_order;
+```
+
+Relaxed allows results to be slightly out of order by distance, but provides better recall
+
+```sql
+SET hnsw.iterative_scan = relaxed_order;
+# or
+SET ivfflat.iterative_scan = relaxed_order;
+```
+
+With relaxed ordering, you can use a [materialized CTE](https://www.postgresql.org/docs/current/queries-with.html#QUERIES-WITH-CTE-MATERIALIZATION) to get strict ordering
+
+```sql
+WITH relaxed_results AS MATERIALIZED (
+    SELECT id, embedding <-> '[1,2,3]' AS distance FROM items WHERE category_id = 123 ORDER BY distance LIMIT 5
+) SELECT * FROM relaxed_results ORDER BY distance;
+```
+
+For queries that filter by distance, use a materialized CTE and place the distance filter outside of it for best performance (due to the [current behavior](https://www.postgresql.org/message-id/flat/CAOdR5yGUoMQ6j7M5hNUXrySzaqZVGf_Ne%2B8fwZMRKTFxU1nbJg%40mail.gmail.com) of the Postgres executor)
+
+```sql
+WITH nearest_results AS MATERIALIZED (
+    SELECT id, embedding <-> '[1,2,3]' AS distance FROM items ORDER BY distance LIMIT 5
+) SELECT * FROM nearest_results WHERE distance < 5 ORDER BY distance;
+```
+
+Note: Place any other filters inside the CTE
+
+### Iterative Scan Options
+
+Since scanning a large portion of an approximate index is expensive, there are options to control when a scan ends.
+
+#### HNSW
+
+Specify the max number of tuples to visit (20,000 by default)
+
+```sql
+SET hnsw.max_scan_tuples = 20000;
+```
+
+Note: This is approximate and does not affect the initial scan
+
+Specify the max amount of memory to use, as a multiple of `work_mem` (1 by default)
+
+```sql
+SET hnsw.scan_mem_multiplier = 2;
+```
+
+Note: Try increasing this if increasing `hnsw.max_scan_tuples` does not improve recall
+
+#### IVFFlat
+
+Specify the max number of probes
+
+```sql
+SET ivfflat.max_probes = 100;
+```
+
+Note: If this is lower than `ivfflat.probes`, `ivfflat.probes` will be used
+
 ## Half-Precision Vectors
 
 *Added in 0.7.0*
@@ -1057,7 +1153,7 @@ This adds pgvector to the [Postgres image](https://hub.docker.com/_/postgres) (r
 You can also build the image manually:
 
 ```sh
-git clone --branch v0.7.4 https://github.com/pgvector/pgvector.git
+git clone --branch v0.8.0 https://github.com/pgvector/pgvector.git
 cd pgvector
 docker build --pull --build-arg PG_MAJOR=17 -t myuser/pgvector .
 ```

src/hnsw.c

@@ -18,16 +18,17 @@
 #define MarkGUCPrefixReserved(x) EmitWarningsOnPlaceholders(x)
 #endif
 
-static const struct config_enum_entry hnsw_iterative_search_options[] = {
-	{"off", HNSW_ITERATIVE_SEARCH_OFF, false},
-	{"relaxed_order", HNSW_ITERATIVE_SEARCH_RELAXED, false},
-	{"strict_order", HNSW_ITERATIVE_SEARCH_STRICT, false},
+static const struct config_enum_entry hnsw_iterative_scan_options[] = {
+	{"off", HNSW_ITERATIVE_SCAN_OFF, false},
+	{"relaxed_order", HNSW_ITERATIVE_SCAN_RELAXED, false},
+	{"strict_order", HNSW_ITERATIVE_SCAN_STRICT, false},
 	{NULL, 0, false}
 };
 
 int			hnsw_ef_search;
-int			hnsw_max_search_tuples;
-int			hnsw_iterative_search;
+int			hnsw_iterative_scan;
+int			hnsw_max_scan_tuples;
+double		hnsw_scan_mem_multiplier;
 int			hnsw_lock_tranche_id;
 static relopt_kind hnsw_relopt_kind;
 
@@ -78,14 +79,19 @@ HnswInit(void)
 							"Valid range is 1..1000.", &hnsw_ef_search,
 							HNSW_DEFAULT_EF_SEARCH, HNSW_MIN_EF_SEARCH, HNSW_MAX_EF_SEARCH, PGC_USERSET, 0, NULL, NULL, NULL);
 
-	DefineCustomEnumVariable("hnsw.iterative_search", "Sets the iterative search mode",
-							 NULL, &hnsw_iterative_search,
-							 HNSW_ITERATIVE_SEARCH_OFF, hnsw_iterative_search_options, PGC_USERSET, 0, NULL, NULL, NULL);
+	DefineCustomEnumVariable("hnsw.iterative_scan", "Sets the mode for iterative scans",
+							 NULL, &hnsw_iterative_scan,
+							 HNSW_ITERATIVE_SCAN_OFF, hnsw_iterative_scan_options, PGC_USERSET, 0, NULL, NULL, NULL);
 
-	/* This is approximate and does not apply to the initial scan */
-	DefineCustomIntVariable("hnsw.max_search_tuples", "Sets the max number of candidates to visit for iterative search",
-							"-1 means no limit", &hnsw_max_search_tuples,
-							-1, -1, INT_MAX, PGC_USERSET, 0, NULL, NULL, NULL);
+	/* This is approximate and does not affect the initial scan */
+	DefineCustomIntVariable("hnsw.max_scan_tuples", "Sets the max number of tuples to visit for iterative scans",
+							NULL, &hnsw_max_scan_tuples,
+							20000, 1, INT_MAX, PGC_USERSET, 0, NULL, NULL, NULL);
+
+	/* Same range as hash_mem_multiplier */
+	DefineCustomRealVariable("hnsw.scan_mem_multiplier", "Sets the multiple of work_mem to use for iterative scans",
+							 NULL, &hnsw_scan_mem_multiplier,
+							 1, 1, 1000, PGC_USERSET, 0, NULL, NULL, NULL);
 
 	MarkGUCPrefixReserved("hnsw");
 }
@@ -131,6 +137,10 @@ hnswcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 		*indexSelectivity = 0;
 		*indexCorrelation = 0;
 		*indexPages = 0;
+#if PG_VERSION_NUM >= 180000
+		/* See "On disable_cost" thread on pgsql-hackers */
+		path->path.disabled_nodes = 2;
+#endif
 		return;
 	}
 

src/hnsw.h

@@ -109,16 +109,17 @@
 
 /* Variables */
 extern int	hnsw_ef_search;
-extern int	hnsw_iterative_search;
-extern int	hnsw_max_search_tuples;
+extern int	hnsw_iterative_scan;
+extern int	hnsw_max_scan_tuples;
+extern double hnsw_scan_mem_multiplier;
 extern int	hnsw_lock_tranche_id;
 
-typedef enum HnswIterativeSearchMode
+typedef enum HnswIterativeScanMode
 {
-	HNSW_ITERATIVE_SEARCH_OFF,
-	HNSW_ITERATIVE_SEARCH_RELAXED,
-	HNSW_ITERATIVE_SEARCH_STRICT
-}			HnswIterativeSearchMode;
+	HNSW_ITERATIVE_SCAN_OFF,
+	HNSW_ITERATIVE_SCAN_RELAXED,
+	HNSW_ITERATIVE_SCAN_STRICT
+}			HnswIterativeScanMode;
 
 typedef struct HnswElementData HnswElementData;
 typedef struct HnswNeighborArray HnswNeighborArray;
@@ -372,6 +373,7 @@ typedef struct HnswScanOpaqueData
 	int			m;
 	int64		tuples;
 	double		previousDistance;
+	Size		maxMemory;
 	MemoryContext tmpCtx;
 
 	/* Support functions */

src/hnswscan.c

@@ -41,7 +41,7 @@ GetScanItems(IndexScanDesc scan, Datum value)
 		ep = w;
 	}
 
-	return HnswSearchLayer(base, q, ep, hnsw_ef_search, 0, index, support, m, false, NULL, &so->v, hnsw_iterative_search != HNSW_ITERATIVE_SEARCH_OFF ? &so->discarded : NULL, true, &so->tuples);
+	return HnswSearchLayer(base, q, ep, hnsw_ef_search, 0, index, support, m, false, NULL, &so->v, hnsw_iterative_scan != HNSW_ITERATIVE_SCAN_OFF ? &so->discarded : NULL, true, &so->tuples);
 }
 
 /*
@@ -121,14 +121,15 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
 {
 	IndexScanDesc scan;
 	HnswScanOpaque so;
+	double		maxMemory;
 
 	scan = RelationGetIndexScan(index, nkeys, norderbys);
 
 	so = (HnswScanOpaque) palloc(sizeof(HnswScanOpaqueData));
 	so->typeInfo = HnswGetTypeInfo(index);
-	so->first = true;
-	so->v.tids = NULL;
-	so->discarded = NULL;
+
+	/* Set support functions */
+	HnswInitSupport(&so->support, index);
 
 	/*
 	 * Use a lower max allocation size than default to allow scanning more
@@ -136,10 +137,12 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
 	 */
 	so->tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
 									   "Hnsw scan temporary context",
-									   0, 8 * 1024, 512 * 1024);
+									   0, 8 * 1024, 256 * 1024);
 
-	/* Set support functions */
-	HnswInitSupport(&so->support, index);
+	/* Calculate max memory */
+	/* Add 256 extra bytes to fill last block when close */
+	maxMemory = (double) work_mem * hnsw_scan_mem_multiplier * 1024.0 + 256;
+	so->maxMemory = Min(maxMemory, (double) SIZE_MAX);
 
 	scan->opaque = so;
 
@@ -154,13 +157,10 @@ hnswrescan(IndexScanDesc scan, ScanKey keys, int nkeys, ScanKey orderbys, int no
 {
 	HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
 
-	if (so->v.tids != NULL)
-		tidhash_reset(so->v.tids);
-
-	if (so->discarded != NULL)
-		pairingheap_reset(so->discarded);
-
 	so->first = true;
+	/* v and discarded are allocated in tmpCtx */
+	so->v.tids = NULL;
+	so->discarded = NULL;
 	so->tuples = 0;
 	so->previousDistance = -get_float8_infinity();
 	MemoryContextReset(so->tmpCtx);
@@ -233,15 +233,15 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 
 		if (list_length(so->w) == 0)
 		{
-			if (hnsw_iterative_search == HNSW_ITERATIVE_SEARCH_OFF)
+			if (hnsw_iterative_scan == HNSW_ITERATIVE_SCAN_OFF)
 				break;
 
 			/* Empty index */
 			if (so->discarded == NULL)
 				break;
 
-			/* Reached max number of tuples */
-			if (hnsw_max_search_tuples != -1 && so->tuples >= hnsw_max_search_tuples)
+			/* Reached max number of tuples or memory limit */
+			if (so->tuples >= hnsw_max_scan_tuples || MemoryContextMemAllocated(so->tmpCtx, false) > so->maxMemory)
 			{
 				if (pairingheap_is_empty(so->discarded))
 					break;
@@ -249,21 +249,6 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 				/* Return remaining tuples */
 				so->w = lappend(so->w, HnswGetSearchCandidate(w_node, pairingheap_remove_first(so->discarded)));
 			}
-			/* Prevent scans from consuming too much memory */
-			else if (MemoryContextMemAllocated(so->tmpCtx, false) > (Size) work_mem * 1024L)
-			{
-				if (pairingheap_is_empty(so->discarded))
-				{
-					ereport(DEBUG1,
-							(errmsg("hnsw index scan exceeded work_mem after " INT64_FORMAT " tuples", so->tuples),
-							 errhint("Increase work_mem to scan more tuples.")));
-
-					break;
-				}
-
-				/* Return remaining tuples */
-				so->w = lappend(so->w, HnswGetSearchCandidate(w_node, pairingheap_remove_first(so->discarded)));
-			}
 			else
 			{
 				/*
@@ -299,7 +284,7 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 			so->w = list_delete_last(so->w);
 
 			/* Mark memory as free for next iteration */
-			if (hnsw_iterative_search != HNSW_ITERATIVE_SEARCH_OFF)
+			if (hnsw_iterative_scan != HNSW_ITERATIVE_SCAN_OFF)
 			{
 				pfree(element);
 				pfree(sc);
@@ -310,7 +295,7 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
 
 		heaptid = &element->heaptids[--element->heaptidsLength];
 
-		if (hnsw_iterative_search == HNSW_ITERATIVE_SEARCH_STRICT)
+		if (hnsw_iterative_scan == HNSW_ITERATIVE_SCAN_STRICT)
 		{
 			if (sc->distance < so->previousDistance)
 				continue;

src/hnswutils.c

@@ -857,6 +857,7 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
 		{
 			AddToVisited(base, v, sc->element, inMemory, &found);
 
+			/* OK to count elements instead of tuples */
 			if (tuples != NULL)
 				(*tuples)++;
 		}
@@ -889,6 +890,7 @@ HnswSearchLayer(char *base, HnswQuery * q, List *ep, int ef, int lc, Relation in
 		else
 			HnswLoadUnvisitedFromDisk(cElement, unvisited, &unvisitedLength, v, index, m, lm, lc);
 
+		/* OK to count elements instead of tuples */
 		if (tuples != NULL)
 			(*tuples) += unvisitedLength;
 

src/ivfflat.c

@@ -17,13 +17,13 @@
 #endif
 
 int			ivfflat_probes;
-int			ivfflat_iterative_search;
+int			ivfflat_iterative_scan;
 int			ivfflat_max_probes;
 static relopt_kind ivfflat_relopt_kind;
 
-static const struct config_enum_entry ivfflat_iterative_search_options[] = {
-	{"off", IVFFLAT_ITERATIVE_SEARCH_OFF, false},
-	{"relaxed_order", IVFFLAT_ITERATIVE_SEARCH_RELAXED, false},
+static const struct config_enum_entry ivfflat_iterative_scan_options[] = {
+	{"off", IVFFLAT_ITERATIVE_SCAN_OFF, false},
+	{"relaxed_order", IVFFLAT_ITERATIVE_SCAN_RELAXED, false},
 	{NULL, 0, false}
 };
 
@@ -41,14 +41,14 @@ IvfflatInit(void)
 							"Valid range is 1..lists.", &ivfflat_probes,
 							IVFFLAT_DEFAULT_PROBES, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, 0, NULL, NULL, NULL);
 
-	DefineCustomEnumVariable("ivfflat.iterative_search", "Sets the iterative search mode",
-							 NULL, &ivfflat_iterative_search,
-							 IVFFLAT_ITERATIVE_SEARCH_OFF, ivfflat_iterative_search_options, PGC_USERSET, 0, NULL, NULL, NULL);
+	DefineCustomEnumVariable("ivfflat.iterative_scan", "Sets the mode for iterative scans",
+							 NULL, &ivfflat_iterative_scan,
+							 IVFFLAT_ITERATIVE_SCAN_OFF, ivfflat_iterative_scan_options, PGC_USERSET, 0, NULL, NULL, NULL);
 
 	/* If this is less than probes, probes is used */
-	DefineCustomIntVariable("ivfflat.max_probes", "Sets the max number of probes for iterative search",
-							"-1 means no limit", &ivfflat_max_probes,
-							-1, -1, IVFFLAT_MAX_LISTS, PGC_USERSET, 0, NULL, NULL, NULL);
+	DefineCustomIntVariable("ivfflat.max_probes", "Sets the max number of probes for iterative scans",
+							NULL, &ivfflat_max_probes,
+							IVFFLAT_MAX_LISTS, IVFFLAT_MIN_LISTS, IVFFLAT_MAX_LISTS, PGC_USERSET, 0, NULL, NULL, NULL);
 
 	MarkGUCPrefixReserved("ivfflat");
 }
@@ -99,6 +99,10 @@ ivfflatcostestimate(PlannerInfo *root, IndexPath *path, double loop_count,
 		*indexSelectivity = 0;
 		*indexCorrelation = 0;
 		*indexPages = 0;
+#if PG_VERSION_NUM >= 180000
+		/* See "On disable_cost" thread on pgsql-hackers */
+		path->path.disabled_nodes = 2;
+#endif
 		return;
 	}
 

src/ivfflat.h

@@ -80,14 +80,14 @@
 
 /* Variables */
 extern int	ivfflat_probes;
-extern int	ivfflat_iterative_search;
+extern int	ivfflat_iterative_scan;
 extern int	ivfflat_max_probes;
 
-typedef enum IvfflatIterativeSearchMode
+typedef enum IvfflatIterativeScanMode
 {
-	IVFFLAT_ITERATIVE_SEARCH_OFF,
-	IVFFLAT_ITERATIVE_SEARCH_RELAXED
-}			IvfflatIterativeSearchMode;
+	IVFFLAT_ITERATIVE_SCAN_OFF,
+	IVFFLAT_ITERATIVE_SCAN_RELAXED
+}			IvfflatIterativeScanMode;
 
 typedef struct VectorArrayData
 {

src/ivfscan.c

@@ -114,7 +114,6 @@ GetScanItems(IndexScanDesc scan, Datum value)
 {
 	IvfflatScanOpaque so = (IvfflatScanOpaque) scan->opaque;
 	TupleDesc	tupdesc = RelationGetDescr(scan->indexRelation);
-	double		tuples = 0;
 	TupleTableSlot *slot = so->vslot;
 	int			batchProbes = 0;
 
@@ -161,8 +160,6 @@ GetScanItems(IndexScanDesc scan, Datum value)
 				ExecStoreVirtualTuple(slot);
 
 				tuplesort_puttupleslot(so->sortstate, slot);
-
-				tuples++;
 			}
 
 			searchPage = IvfflatPageGetOpaque(page)->nextblkno;
@@ -171,12 +168,6 @@ GetScanItems(IndexScanDesc scan, Datum value)
 		}
 	}
 
-	if (tuples < 100 && ivfflat_iterative_search == IVFFLAT_ITERATIVE_SEARCH_OFF)
-		ereport(DEBUG1,
-				(errmsg("index scan found few tuples"),
-				 errdetail("Index may have been created with little data."),
-				 errhint("Recreate the index and possibly decrease lists.")));
-
 	tuplesort_performsort(so->sortstate);
 
 #if defined(IVFFLAT_MEMORY)
@@ -263,18 +254,8 @@ ivfflatbeginscan(Relation index, int nkeys, int norderbys)
 	/* Get lists and dimensions from metapage */
 	IvfflatGetMetaPageInfo(index, &lists, &dimensions);
 
-	if (ivfflat_iterative_search != IVFFLAT_ITERATIVE_SEARCH_OFF)
-	{
-		maxProbes = ivfflat_max_probes;
-
-		if (maxProbes < 0)
-			maxProbes = lists;
-		else if (maxProbes < probes)
-		{
-			/* TODO Show notice */
-			maxProbes = probes;
-		}
-	}
+	if (ivfflat_iterative_scan != IVFFLAT_ITERATIVE_SCAN_OFF)
+		maxProbes = Max(ivfflat_max_probes, probes);
 	else
 		maxProbes = probes;
 

src/sparsevec.c

@@ -4,6 +4,7 @@
 #include <math.h>
 
 #include "catalog/pg_type.h"
+#include "common/shortest_dec.h"
 #include "common/string.h"
 #include "fmgr.h"
 #include "halfutils.h"
@@ -12,17 +13,10 @@
 #include "sparsevec.h"
 #include "utils/array.h"
 #include "utils/builtins.h"
+#include "utils/float.h"
 #include "utils/lsyscache.h"
 #include "vector.h"
 
-#if PG_VERSION_NUM >= 120000
-#include "common/shortest_dec.h"
-#include "utils/float.h"
-#else
-#include <float.h>
-#include "utils/builtins.h"
-#endif
-
 typedef struct SparseInputElement
 {
 	int32		index;

test/expected/hnsw_vector.out

@@ -104,7 +104,7 @@ DROP TABLE t;
 CREATE TABLE t (val vector(3));
 INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
 CREATE INDEX ON t USING hnsw (val vector_l2_ops);
-SET hnsw.iterative_search = strict_order;
+SET hnsw.iterative_scan = strict_order;
 SET hnsw.ef_search = 1;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
    val   
@@ -114,7 +114,7 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [0,0,0]
 (3 rows)
 
-SET hnsw.iterative_search = relaxed_order;
+SET hnsw.iterative_scan = relaxed_order;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
    val   
 ---------
@@ -123,7 +123,7 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [0,0,0]
 (3 rows)
 
-RESET hnsw.iterative_search;
+RESET hnsw.iterative_scan;
 RESET hnsw.ef_search;
 DROP TABLE t;
 -- unlogged
@@ -165,21 +165,31 @@ SET hnsw.ef_search = 0;
 ERROR:  0 is outside the valid range for parameter "hnsw.ef_search" (1 .. 1000)
 SET hnsw.ef_search = 1001;
 ERROR:  1001 is outside the valid range for parameter "hnsw.ef_search" (1 .. 1000)
-SHOW hnsw.iterative_search;
- hnsw.iterative_search 
------------------------
+SHOW hnsw.iterative_scan;
+ hnsw.iterative_scan 
+---------------------
  off
 (1 row)
 
-SET hnsw.iterative_search = on;
-ERROR:  invalid value for parameter "hnsw.iterative_search": "on"
+SET hnsw.iterative_scan = on;
+ERROR:  invalid value for parameter "hnsw.iterative_scan": "on"
 HINT:  Available values: off, relaxed_order, strict_order.
-SHOW hnsw.max_search_tuples;
- hnsw.max_search_tuples 
-------------------------
- -1
+SHOW hnsw.max_scan_tuples;
+ hnsw.max_scan_tuples 
+----------------------
+ 20000
 (1 row)
 
-SET hnsw.max_search_tuples = -2;
-ERROR:  -2 is outside the valid range for parameter "hnsw.max_search_tuples" (-1 .. 2147483647)
+SET hnsw.max_scan_tuples = 0;
+ERROR:  0 is outside the valid range for parameter "hnsw.max_scan_tuples" (1 .. 2147483647)
+SHOW hnsw.scan_mem_multiplier;
+ hnsw.scan_mem_multiplier 
+--------------------------
+ 1
+(1 row)
+
+SET hnsw.scan_mem_multiplier = 0;
+ERROR:  0 is outside the valid range for parameter "hnsw.scan_mem_multiplier" (1 .. 1000)
+SET hnsw.scan_mem_multiplier = 1001;
+ERROR:  1001 is outside the valid range for parameter "hnsw.scan_mem_multiplier" (1 .. 1000)
 DROP TABLE t;

test/expected/ivfflat_vector.out

@@ -86,7 +86,7 @@ DROP TABLE t;
 CREATE TABLE t (val vector(3));
 INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
 CREATE INDEX ON t USING ivfflat (val vector_l2_ops) WITH (lists = 3);
-SET ivfflat.iterative_search = relaxed_order;
+SET ivfflat.iterative_scan = relaxed_order;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
    val   
 ---------
@@ -95,13 +95,6 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [0,0,0]
 (3 rows)
 
-SET ivfflat.max_probes = 0;
-SELECT * FROM t ORDER BY val <-> '[3,3,3]';
-   val   
----------
- [1,2,3]
-(1 row)
-
 SET ivfflat.max_probes = 1;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
    val   
@@ -117,7 +110,7 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
  [1,1,1]
 (2 rows)
 
-RESET ivfflat.iterative_search;
+RESET ivfflat.iterative_scan;
 RESET ivfflat.max_probes;
 DROP TABLE t;
 -- unlogged
@@ -151,23 +144,23 @@ SET ivfflat.probes = 0;
 ERROR:  0 is outside the valid range for parameter "ivfflat.probes" (1 .. 32768)
 SET ivfflat.probes = 32769;
 ERROR:  32769 is outside the valid range for parameter "ivfflat.probes" (1 .. 32768)
-SHOW ivfflat.iterative_search;
- ivfflat.iterative_search 
---------------------------
+SHOW ivfflat.iterative_scan;
+ ivfflat.iterative_scan 
+------------------------
  off
 (1 row)
 
-SET ivfflat.iterative_search = on;
-ERROR:  invalid value for parameter "ivfflat.iterative_search": "on"
+SET ivfflat.iterative_scan = on;
+ERROR:  invalid value for parameter "ivfflat.iterative_scan": "on"
 HINT:  Available values: off, relaxed_order.
 SHOW ivfflat.max_probes;
  ivfflat.max_probes 
 --------------------
- -1
+ 32768
 (1 row)
 
-SET ivfflat.max_probes = -2;
-ERROR:  -2 is outside the valid range for parameter "ivfflat.max_probes" (-1 .. 32768)
+SET ivfflat.max_probes = 0;
+ERROR:  0 is outside the valid range for parameter "ivfflat.max_probes" (1 .. 32768)
 SET ivfflat.max_probes = 32769;
-ERROR:  32769 is outside the valid range for parameter "ivfflat.max_probes" (-1 .. 32768)
+ERROR:  32769 is outside the valid range for parameter "ivfflat.max_probes" (1 .. 32768)
 DROP TABLE t;

test/sql/hnsw_vector.sql

@@ -63,14 +63,14 @@ CREATE TABLE t (val vector(3));
 INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
 CREATE INDEX ON t USING hnsw (val vector_l2_ops);
 
-SET hnsw.iterative_search = strict_order;
+SET hnsw.iterative_scan = strict_order;
 SET hnsw.ef_search = 1;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
-SET hnsw.iterative_search = relaxed_order;
+SET hnsw.iterative_scan = relaxed_order;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
-RESET hnsw.iterative_search;
+RESET hnsw.iterative_scan;
 RESET hnsw.ef_search;
 DROP TABLE t;
 
@@ -98,12 +98,17 @@ SHOW hnsw.ef_search;
 SET hnsw.ef_search = 0;
 SET hnsw.ef_search = 1001;
 
-SHOW hnsw.iterative_search;
+SHOW hnsw.iterative_scan;
 
-SET hnsw.iterative_search = on;
+SET hnsw.iterative_scan = on;
 
-SHOW hnsw.max_search_tuples;
+SHOW hnsw.max_scan_tuples;
 
-SET hnsw.max_search_tuples = -2;
+SET hnsw.max_scan_tuples = 0;
+
+SHOW hnsw.scan_mem_multiplier;
+
+SET hnsw.scan_mem_multiplier = 0;
+SET hnsw.scan_mem_multiplier = 1001;
 
 DROP TABLE t;

test/sql/ivfflat_vector.sql

@@ -50,10 +50,7 @@ CREATE TABLE t (val vector(3));
 INSERT INTO t (val) VALUES ('[0,0,0]'), ('[1,2,3]'), ('[1,1,1]'), (NULL);
 CREATE INDEX ON t USING ivfflat (val vector_l2_ops) WITH (lists = 3);
 
-SET ivfflat.iterative_search = relaxed_order;
-SELECT * FROM t ORDER BY val <-> '[3,3,3]';
-
-SET ivfflat.max_probes = 0;
+SET ivfflat.iterative_scan = relaxed_order;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
 SET ivfflat.max_probes = 1;
@@ -62,7 +59,7 @@ SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 SET ivfflat.max_probes = 2;
 SELECT * FROM t ORDER BY val <-> '[3,3,3]';
 
-RESET ivfflat.iterative_search;
+RESET ivfflat.iterative_scan;
 RESET ivfflat.max_probes;
 DROP TABLE t;
 
@@ -87,13 +84,13 @@ SHOW ivfflat.probes;
 SET ivfflat.probes = 0;
 SET ivfflat.probes = 32769;
 
-SHOW ivfflat.iterative_search;
+SHOW ivfflat.iterative_scan;
 
-SET ivfflat.iterative_search = on;
+SET ivfflat.iterative_scan = on;
 
 SHOW ivfflat.max_probes;
 
-SET ivfflat.max_probes = -2;
+SET ivfflat.max_probes = 0;
 SET ivfflat.max_probes = 32769;
 
 DROP TABLE t;

test/t/041_ivfflat_iterative_scan.pl

@@ -23,7 +23,7 @@ $node->safe_psql("postgres", "CREATE INDEX ON tst USING ivfflat (v vector_l2_ops
 my $count = $node->safe_psql("postgres", qq(
 	SET enable_seqscan = off;
 	SET ivfflat.probes = 10;
-	SET ivfflat.iterative_search = relaxed_order;
+	SET ivfflat.iterative_scan = relaxed_order;
 	SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst LIMIT 1) LIMIT 11) t;
 ));
 is($count, 10);
@@ -39,7 +39,7 @@ foreach ((30, 50, 70))
 		$count = $node->safe_psql("postgres", qq(
 			SET enable_seqscan = off;
 			SET ivfflat.probes = 10;
-			SET ivfflat.iterative_search = relaxed_order;
+			SET ivfflat.iterative_scan = relaxed_order;
 			SET ivfflat.max_probes = $max_probes;
 			SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst WHERE i = $i) LIMIT 11) t;
 		));

test/t/042_ivfflat_iterative_scan_recall.pl

@@ -19,7 +19,7 @@ sub test_recall
 	my $explain = $node->safe_psql("postgres", qq(
 		SET enable_seqscan = off;
 		SET ivfflat.probes = $probes;
-		SET ivfflat.iterative_search = relaxed_order;
+		SET ivfflat.iterative_scan = relaxed_order;
 		EXPLAIN ANALYZE SELECT i FROM tst WHERE i % $c = 0 ORDER BY v $operator '$queries[0]' LIMIT $limit;
 	));
 	like($explain, qr/Index Scan using idx on tst/);
@@ -29,7 +29,7 @@ sub test_recall
 		my $actual = $node->safe_psql("postgres", qq(
 			SET enable_seqscan = off;
 			SET ivfflat.probes = $probes;
-			SET ivfflat.iterative_search = relaxed_order;
+			SET ivfflat.iterative_scan = relaxed_order;
 			SELECT i FROM tst WHERE i % $c = 0 ORDER BY v $operator '$queries[$i]' LIMIT $limit;
 		));
 		my @actual_ids = split("\n", $actual);

test/t/043_hnsw_iterative_scan.pl

@@ -26,8 +26,9 @@ $node->safe_psql("postgres", qq(
 
 my $count = $node->safe_psql("postgres", qq(
 	SET enable_seqscan = off;
-	SET hnsw.iterative_search = relaxed_order;
-	SET work_mem = '8MB';
+	SET hnsw.iterative_scan = relaxed_order;
+	SET hnsw.max_scan_tuples = 100000;
+	SET hnsw.scan_mem_multiplier = 2;
 	SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst LIMIT 1) LIMIT 11) t;
 ));
 is($count, 10);
@@ -42,9 +43,9 @@ foreach ((30000, 50000, 70000))
 	{
 		$count = $node->safe_psql("postgres", qq(
 			SET enable_seqscan = off;
-			SET hnsw.iterative_search = relaxed_order;
-			SET hnsw.max_search_tuples = $max_tuples;
-			SET work_mem = '8MB';
+			SET hnsw.iterative_scan = relaxed_order;
+			SET hnsw.max_scan_tuples = $max_tuples;
+			SET hnsw.scan_mem_multiplier = 2;
 			SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst WHERE i = $i) LIMIT 11) t;
 		));
 		$sum += $count;
@@ -55,13 +56,4 @@ foreach ((30000, 50000, 70000))
 	cmp_ok($avg, '<', $expected + 2);
 }
 
-my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
-	SET enable_seqscan = off;
-	SET hnsw.iterative_search = relaxed_order;
-	SET client_min_messages = debug1;
-	SET work_mem = '2MB';
-	SELECT COUNT(*) FROM (SELECT v FROM tst WHERE i % 10000 = 0 ORDER BY v <-> (SELECT v FROM tst LIMIT 1) LIMIT 11) t;
-));
-like($stderr, qr/hnsw index scan exceeded work_mem after \d+ tuples/);
-
 done_testing();

test/t/044_hnsw_iterative_scan_recall.pl

@@ -21,7 +21,7 @@ sub test_recall
 	my $explain = $node->safe_psql("postgres", qq(
 		SET enable_seqscan = off;
 		SET hnsw.ef_search = $ef_search;
-		SET hnsw.iterative_search = $mode;
+		SET hnsw.iterative_scan = $mode;
 		EXPLAIN ANALYZE SELECT i FROM tst WHERE i % $c = 0 ORDER BY v $operator '$queries[0]' LIMIT $limit;
 	));
 	like($explain, qr/Index Scan using idx on tst/);
@@ -31,7 +31,7 @@ sub test_recall
 		my $actual = $node->safe_psql("postgres", qq(
 			SET enable_seqscan = off;
 			SET hnsw.ef_search = $ef_search;
-			SET hnsw.iterative_search = $mode;
+			SET hnsw.iterative_scan = $mode;
 			SELECT i FROM tst WHERE i % $c = 0 ORDER BY v $operator '$queries[$i]' LIMIT $limit;
 		));
 		my @actual_ids = split("\n", $actual);

vector.control

@@ -1,4 +1,4 @@
 comment = 'vector data type and ivfflat and hnsw access methods'
-default_version = '0.7.4'
+default_version = '0.8.0'
 module_pathname = '$libdir/vector'
 relocatable = true