Compare commits

..

100 Commits

Author SHA1 Message Date
Andrew Kane
13d3bae99b Less space [skip ci] 2024-01-22 18:43:26 -08:00
Andrew Kane
3b9c4c55ee Use consistent order [skip ci] 2024-01-22 18:39:52 -08:00
Andrew Kane
06e2052e40 Added InitVisited [skip ci] 2024-01-22 18:37:37 -08:00
Andrew Kane
891072df68 Added base [skip ci] 2024-01-22 18:02:24 -08:00
Andrew Kane
e9a96d341f Removed prefix for static functions 2024-01-22 17:58:42 -08:00
Andrew Kane
a45ca414f4 Keep logic consistent with insert [skip ci] 2024-01-22 17:43:04 -08:00
Andrew Kane
b4f3cc3e13 Moved outside loop [skip ci] 2024-01-22 17:26:26 -08:00
Andrew Kane
c01cf8a315 Renamed ApplyChanges to UpdateGraph [skip ci] 2024-01-22 17:02:02 -08:00
Andrew Kane
3ecb9a3cb2 Renamed HnswInsertElement to HnswFindElementNeighbors [skip ci] 2024-01-22 16:59:08 -08:00
Andrew Kane
a069e18fe4 Improved function names [skip ci] 2024-01-22 16:48:50 -08:00
Andrew Kane
5174a23094 Updated comment [skip ci] 2024-01-22 16:45:42 -08:00
Andrew Kane
16d7de79f6 Improved function names [skip ci] 2024-01-22 16:43:50 -08:00
Andrew Kane
e54ec4d637 Improved code [skip ci] 2024-01-22 16:37:32 -08:00
Andrew Kane
cc641002d3 Updated comments [skip ci] 2024-01-22 10:59:46 -08:00
Heikki Linnakangas
2f9b1e2893 xAdd overview comment on how HNSW build works (#419)
And rewrite some of the comments in InsertTuple(), to also give more
of a high-level overview of the flow.
2024-01-22 10:50:12 -08:00
Andrew Kane
a3e4fbf6aa Use shared lock for copying neighbors to local memory 2024-01-19 13:44:25 -08:00
Andrew Kane
09a4ec29a0 Added InsertTupleInMemory 2024-01-19 01:27:45 -08:00
Heikki Linnakangas
ca3b4cd029 Remove HnswSpool
It was just used to pass heap/index relations to
HnswParallelScanAndInsert. I think it was copied from nbtsort.c, which
is more complicated. I don't think we need a struct like this.

(That said, I actually think that we should have a state object that
would hold fields like 'heap', 'index', 'procinfo', 'collation'
etc. Passing that object around would simplify the signatures of many
functions. But that's a different story).
2024-01-19 00:26:47 -08:00
Heikki Linnakangas
d96e486274 Remove unused 'scantuplesortstates' field 2024-01-19 00:26:47 -08:00
Heikki Linnakangas
88213186a5 Remove unused argument 2024-01-19 00:26:47 -08:00
Andrew Kane
7dd9534894 Use same locking as insert 2024-01-19 00:18:29 -08:00
Andrew Kane
d801a843f4 Removed HnswPtrSetNull to avoid setting relptr_off directly 2024-01-16 17:08:13 -08:00
Andrew Kane
1458c7bb2a Improved code [skip ci] 2024-01-16 14:03:28 -08:00
Andrew Kane
cad48d9203 Improved locking 2024-01-16 13:34:55 -08:00
Heikki Linnakangas
719b4b7436 Use LWLocks instead of SpinLocks (#410)
Spinlocks should be held only for a few instructions, for multiple
reasons:

- You have to be very careful not to elog() out while holding a
  spinlock, because there is no mechanism to release the spinlock on
  error.

- Waiters can waste a lot of cycles spinning if the lock is
  contended. I you wait on a spinlock for too long, the PostgreSQL
  implementation will actually PANIC, see s_lock_stuck().

The flushLock is particularly problematic. It is held in exclusive
mode, which means it holds a spinlock, over the call to
FlushPages(). FlushPages() performs lots of I/O so it can take a very
long time (>= minutes), and can also easily error out for various
reasons.

allocatorLock would perhaps be OK as a spinlocks, but even that feels
a bit heavy, so I converted that to an LWLock, too.

entryLock is usually held for a very short time, in shared mode, so
that would be fine as a spinlock. However, in the rare case that the
entry point is updated, it's held for a very long time. An LWLock used
in shared mode is about as fast a spinlock, that path is pretty
heavily optimized.

I think we have some problems with the per-element spinlocks too. In
HnswUpdateNeighborPagesInMemory(), it's held over a call to
HnswUpdateConnection(), but HnswUpdateConnection() can error out at
least in case of an out-of-memory error (it uses lappend(), which
calls palloc()). It also calls the distance function, and I don't
think they are guaranteed to be ereport-free either. However, I didn't
address that in this PR, it needs a bit more thinking.
2024-01-16 13:25:03 -08:00
Andrew Kane
fa0acbf62d Fixed CI 2024-01-15 19:55:46 -08:00
Andrew Kane
1612b84069 Fixed error on Windows [skip ci] 2024-01-15 19:33:16 -08:00
Andrew Kane
2f9371516d Leave space for other objects in shared memory 2024-01-15 19:17:50 -08:00
Andrew Kane
9d3e4e74df Added support for in-memory parallel index builds for HNSW 2024-01-15 15:07:31 -08:00
Andrew Kane
0ce497a1b1 Updated Homebrew note [skip ci] 2024-01-15 12:12:04 -08:00
Andrew Kane
c7d60346d8 Improved macro [skip ci] 2024-01-13 20:02:41 -08:00
Andrew Kane
597bfdc76b Added HnswGetNeighbors macro 2024-01-13 20:00:34 -08:00
Andrew Kane
cbf3eb4fa5 Improved HNSW build and insert code 2024-01-13 10:07:42 -08:00
Andrew Kane
cacd389f6d Improved pattern for duplicates 2024-01-12 14:30:13 -08:00
Andrew Kane
423cc2b06c Homebrew now adds to postgresql@15 as well [skip ci] 2024-01-11 16:45:50 -08:00
Andrew Kane
85c4ef6a14 Updated Postgres versions in readme [skip ci] 2024-01-11 12:36:24 -08:00
Andrew Kane
c6160a783a Homebrew now adds to postgresql@16 [skip ci] 2024-01-11 12:32:14 -08:00
Andrew Kane
1881b857f9 Simplified code 2024-01-09 18:53:31 -08:00
Andrew Kane
51bde5fb22 Updated readme [skip ci] 2024-01-09 14:38:25 -08:00
Andrew Kane
10e65ce349 Added note about maintenance_work_mem [skip ci] 2024-01-09 14:31:54 -08:00
Andrew Kane
61279f5a59 Updated readme [skip ci] 2024-01-09 14:26:55 -08:00
Andrew Kane
72b3889e26 Updated readme [skip ci] 2024-01-09 14:22:19 -08:00
Andrew Kane
bb21b2decf Updated readme [skip ci] 2024-01-09 14:19:01 -08:00
Andrew Kane
8a65c0e831 Moved section [skip ci] 2024-01-09 13:33:03 -08:00
Andrew Kane
7d75d423e4 Added section on index build time [skip ci] 2024-01-09 13:27:27 -08:00
Andrew Kane
6cad1f5de0 Updated example [skip ci] 2024-01-09 13:04:47 -08:00
Andrew Kane
67eeade63c Moved HNSW first in readme [skip ci] 2024-01-09 13:04:18 -08:00
Andrew Kane
108fb09d7b Improved code [skip ci] 2024-01-08 17:54:49 -08:00
Andrew Kane
65d060ac86 Reverted FlushPages pattern for parallel builds 2024-01-08 10:45:31 -08:00
Andrew Kane
62ee33bb92 Improved locking code 2024-01-08 09:05:12 -08:00
Andrew Kane
520e274dde Improved locking code 2024-01-07 22:34:41 -08:00
Andrew Kane
9e680884bd Moved indtuples to HnswGraph 2024-01-07 22:23:49 -08:00
Andrew Kane
19a0e1b341 Moved graph to separate struct 2024-01-07 20:15:30 -08:00
Andrew Kane
c7fe1571ee Improved code 2024-01-07 18:30:51 -08:00
Andrew Kane
cb4c770df2 Switched to slist for elements to reduce allocations and remove limit 2024-01-07 18:26:19 -08:00
Andrew Kane
85fdecd79b Moved FlushPages before HnswEndParallel 2024-01-07 17:50:46 -08:00
Andrew Kane
6132428914 Improved number of parallel workers for HNSW index builds - closes #397 2024-01-05 19:46:08 -08:00
Andrew Kane
81d13bd40f Improved code [skip ci] 2024-01-03 13:53:23 -05:00
Andrew Kane
8ee37b60a0 Improved memory estimate for HNSW index builds 2024-01-03 13:47:50 -05:00
Andrew Kane
9b73b3d1a6 Reduced memory and allocations for heap TIDs - closes #385 2024-01-03 13:41:34 -05:00
Andrew Kane
cae630784b Improved BuildCallback [skip ci] 2023-12-30 20:55:29 -05:00
Andrew Kane
d87bcd2deb Added comments [skip ci] 2023-12-30 18:29:01 -05:00
Andrew Kane
736576220a Improved BuildCallback 2023-12-30 18:24:03 -05:00
Andrew Kane
a508b120c1 Added IVFFLAT_MEMORY flag to show memory usage [skip ci] 2023-12-24 09:27:09 -05:00
Andrew Kane
9a782d29f8 Use consistent style [skip ci] 2023-12-22 16:41:25 -05:00
Andrew Kane
1e422cd62b Improved readability [skip ci] 2023-12-22 16:39:13 -05:00
Andrew Kane
569c69580a Improved InsertTuple code - #384
Co-authored-by: Heikki Linnakangas <heikki.linnakangas@iki.fi>
2023-12-22 15:08:28 -05:00
Andrew Kane
59509c3a17 Added extra 5% to memory estimate 2023-12-22 14:04:05 -05:00
Andrew Kane
61738846af Updated comment [skip ci] 2023-12-22 14:03:33 -05:00
Andrew Kane
e8c3bf0cef Improved memory tracking for HNSW index builds - #384 2023-12-22 13:35:43 -05:00
Andrew Kane
50d1aed3d8 Improved memory usage logging [skip ci] 2023-12-22 13:09:11 -05:00
Andrew Kane
66e14d2434 Updated indentation [skip ci] 2023-12-22 12:59:50 -05:00
Andrew Kane
42cd4c6833 Fixed call to GenerationContextCreate for Postgres < 15 2023-12-22 12:49:07 -05:00
Andrew Kane
dcbe0b6f0d Reduced memory usage for HNSW index builds - #384
Co-authored-by: Heikki Linnakangas <heikki.linnakangas@iki.fi>
2023-12-22 12:41:47 -05:00
Andrew Kane
f61d4087b5 Slightly improved memory estimation [skip ci] 2023-12-21 10:31:36 -05:00
Andrew Kane
57554e5b46 Added todo [skip ci] 2023-12-20 17:52:31 -05:00
Andrew Kane
6738fa0bd7 Added HNSW_MEMORY flag to show memory usage - #384 [skip ci] 2023-12-20 16:49:16 -05:00
Andrew Kane
9ab10aa674 Fixed CI 2023-12-20 16:29:13 -05:00
Andrew Kane
ec41dfa1d7 Mark meta buffer contents as dirty when not logging 2023-12-20 16:20:15 -05:00
Andrew Kane
43e0b3d9d4 Mark buffer contents as dirty when not logging 2023-12-20 16:16:25 -05:00
Andrew Kane
2bff7ccaa2 Improved memory estimation - #384 [skip ci] 2023-12-20 11:10:27 -05:00
Andrew Kane
e88a425c9b Reduced WAL generation for HNSW index builds - thanks @hlinnaka 2023-12-19 20:37:32 -05:00
Andrew Kane
921427ee03 Replace dynahash hash table in HNSW with simplehash for speed - #378
Co-authored-by: Heikki Linnakangas <heikki.linnakangas@iki.fi>
2023-12-17 11:24:13 -05:00
Andrew Kane
a59aa02dd9 Only show message if flushed due to memory [skip ci] 2023-12-04 15:29:23 -08:00
Andrew Kane
2fef497b7e Fixed check 2023-12-04 15:22:12 -08:00
Andrew Kane
0e19a984fb Updated changelog [skip ci] 2023-12-04 15:15:20 -08:00
Andrew Kane
a156f6c7ae Fixed invalid memory alloc request size error with HNSW - fixes #43 2023-12-04 15:00:32 -08:00
Andrew Kane
c653ac524f Simplified code [skip ci] 2023-12-04 12:19:38 -08:00
Andrew Kane
bf0d56e78e Updated comment [skip ci] 2023-12-03 13:09:57 -08:00
Andrew Kane
4d6739a7af Added Lisp to readme [skip ci] 2023-12-03 10:58:54 -08:00
Andrew Kane
ff744214d0 Added Visual Basic to readme [skip ci] 2023-12-03 00:25:07 -08:00
Andrew Kane
7ca9298163 Updated badge [skip ci] 2023-12-01 15:36:49 -08:00
Andrew Kane
ff3bffd9a8 Moved FAQ [skip ci] 2023-11-29 22:09:09 -08:00
Andrew Kane
014753eb9c Added FAQ about memory [skip ci] 2023-11-29 22:02:27 -08:00
Andrew Kane
6763661d3d Added OCaml to readme [skip ci] 2023-11-28 21:42:30 -08:00
Andrew Kane
d287921d15 Added F# to readme [skip ci] 2023-11-28 01:22:12 -08:00
Andrew Kane
5b12ae8225 Added note about ef_construction [skip ci] 2023-11-16 18:53:48 -08:00
Japin Li
4549e8aeb1 Fix coredump about HnswFreeElement() (#357)
The HnswInitElement() allocate an element with not initialize value
filed, which may has garbage that lead HnswFreeElement() free an
invalid pointer.
2023-11-15 16:19:59 -08:00
Andrew Kane
3263b350f5 Updated HnswLoadElementFromTuple to be less vector-specific 2023-11-11 21:14:12 -08:00
Andrew Kane
dfee5d4045 Added support for on-disk parallel index builds for HNSW 2023-11-11 19:29:45 -08:00
17 changed files with 1855 additions and 995 deletions

View File

@@ -62,7 +62,7 @@ jobs:
wget -q https://github.com/postgres/postgres/archive/refs/tags/REL_14_5.tar.gz
tar xf REL_14_5.tar.gz
- run: make prove_installcheck PROVE_FLAGS="-I ./postgres-REL_14_5/src/test/perl" PERL5LIB="/Users/runner/perl5/lib/perl5"
- run: make clean && /usr/local/opt/llvm@15/bin/scan-build --status-bugs make
- run: make clean && /usr/local/opt/llvm@15/bin/scan-build --status-bugs make PG_CFLAGS="-DUSE_ASSERT_CHECKING"
windows:
runs-on: windows-latest
if: ${{ !startsWith(github.ref_name, 'mac') }}

View File

@@ -1,6 +1,10 @@
## 0.6.0 (unreleased)
## 0.5.2 (unreleased)
- Added support for inline filtering with HNSW
- Improved performance of HNSW
- Added support for parallel index builds for HNSW
- Reduced memory usage for HNSW index builds
- Reduced WAL generation for HNSW index builds
- Fixed `invalid memory alloc request size` error with HNSW index build
## 0.5.1 (2023-10-10)

200
README.md
View File

@@ -10,7 +10,7 @@ Store your vectors with the rest of your data. Supports:
Plus [ACID](https://en.wikipedia.org/wiki/ACID) compliance, point-in-time recovery, JOINs, and all of the other [great features](https://www.postgresql.org/about/) of Postgres
[![Build Status](https://github.com/pgvector/pgvector/workflows/build/badge.svg?branch=master)](https://github.com/pgvector/pgvector/actions)
[![Build Status](https://github.com/pgvector/pgvector/actions/workflows/build.yml/badge.svg)](https://github.com/pgvector/pgvector/actions)
## Installation
@@ -161,8 +161,97 @@ You can add an index to use approximate nearest neighbor search, which trades so
Supported index types are:
- [IVFFlat](#ivfflat)
- [HNSW](#hnsw) - added in 0.5.0
- [IVFFlat](#ivfflat)
## HNSW
An HNSW index creates a multilayer graph. It has better query performance than IVFFlat (in terms of speed-recall tradeoff), but has slower build times and uses more memory. Also, an index can be created without any data in the table since there isnt a training step like IVFFlat.
Add an index for each distance function you want to use.
L2 distance
```sql
CREATE INDEX ON items USING hnsw (embedding vector_l2_ops);
```
Inner product
```sql
CREATE INDEX ON items USING hnsw (embedding vector_ip_ops);
```
Cosine distance
```sql
CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);
```
Vectors with up to 2,000 dimensions can be indexed.
### Index Options
Specify HNSW parameters
- `m` - the max number of connections per layer (16 by default)
- `ef_construction` - the size of the dynamic candidate list for constructing the graph (64 by default)
```sql
CREATE INDEX ON items USING hnsw (embedding vector_l2_ops) WITH (m = 16, ef_construction = 64);
```
A higher value of `ef_construction` provides better recall at the cost of index build time / insert speed.
### Query Options
Specify the size of the dynamic candidate list for search (40 by default)
```sql
SET hnsw.ef_search = 100;
```
A higher value provides better recall at the cost of speed.
Use `SET LOCAL` inside a transaction to set it for a single query
```sql
BEGIN;
SET LOCAL hnsw.ef_search = 100;
SELECT ...
COMMIT;
```
### Index Build Time
Indexes build significantly faster when the graph fits into `maintenance_work_mem`
```sql
SET maintenance_work_mem = '8GB';
```
A notice is shown when the graph no longer fits
```text
NOTICE: hnsw graph no longer fits into maintenance_work_mem after 100000 tuples
DETAIL: Building will take significantly more time.
HINT: Increase maintenance_work_mem to speed up builds.
```
Note: Do not set `maintenance_work_mem` so high that it exhausts the memory on the server
### Indexing Progress
Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
```sql
SELECT phase, round(100.0 * blocks_done / nullif(blocks_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
```
The phases for HNSW are:
1. `initializing`
2. `loading tuples`
## IVFFlat
@@ -232,75 +321,6 @@ The phases for IVFFlat are:
Note: `%` is only populated during the `loading tuples` phase
## HNSW
An HNSW index creates a multilayer graph. It has slower build times and uses more memory than IVFFlat, but has better query performance (in terms of speed-recall tradeoff). Theres no training step like IVFFlat, so the index can be created without any data in the table.
Add an index for each distance function you want to use.
L2 distance
```sql
CREATE INDEX ON items USING hnsw (embedding vector_l2_ops);
```
Inner product
```sql
CREATE INDEX ON items USING hnsw (embedding vector_ip_ops);
```
Cosine distance
```sql
CREATE INDEX ON items USING hnsw (embedding vector_cosine_ops);
```
Vectors with up to 2,000 dimensions can be indexed.
### Index Options
Specify HNSW parameters
- `m` - the max number of connections per layer (16 by default)
- `ef_construction` - the size of the dynamic candidate list for constructing the graph (64 by default)
```sql
CREATE INDEX ON items USING hnsw (embedding vector_l2_ops) WITH (m = 16, ef_construction = 64);
```
### Query Options
Specify the size of the dynamic candidate list for search (40 by default)
```sql
SET hnsw.ef_search = 100;
```
A higher value provides better recall at the cost of speed.
Use `SET LOCAL` inside a transaction to set it for a single query
```sql
BEGIN;
SET LOCAL hnsw.ef_search = 100;
SELECT ...
COMMIT;
```
### Indexing Progress
Check [indexing progress](https://www.postgresql.org/docs/current/progress-reporting.html#CREATE-INDEX-PROGRESS-REPORTING) with Postgres 12+
```sql
SELECT phase, round(100.0 * blocks_done / nullif(blocks_total, 0), 1) AS "%" FROM pg_stat_progress_create_index;
```
The phases for HNSW are:
1. `initializing`
2. `loading tuples`
## Filtering
There are a few ways to index nearest neighbor queries with a `WHERE` clause
@@ -315,17 +335,10 @@ Create an index on one [or more](https://www.postgresql.org/docs/current/indexes
CREATE INDEX ON items (category_id);
```
Or a composite HNSW index for approximate search (added in 0.6.0)
```sql
CREATE INDEX ON items USING hnsw (embedding vector_l2_ops, category_id);
```
Or a [partial index](https://www.postgresql.org/docs/current/indexes-partial.html) on the vector column for approximate search
```sql
CREATE INDEX ON items USING ivfflat (embedding vector_l2_ops) WITH (lists = 100)
WHERE (category_id = 123);
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
@@ -383,7 +396,7 @@ Language | Libraries / Examples
--- | ---
C | [pgvector-c](https://github.com/pgvector/pgvector-c)
C++ | [pgvector-cpp](https://github.com/pgvector/pgvector-cpp)
C# | [pgvector-dotnet](https://github.com/pgvector/pgvector-dotnet)
C#, F#, Visual Basic | [pgvector-dotnet](https://github.com/pgvector/pgvector-dotnet)
Crystal | [pgvector-crystal](https://github.com/pgvector/pgvector-crystal)
Dart | [pgvector-dart](https://github.com/pgvector/pgvector-dart)
Elixir | [pgvector-elixir](https://github.com/pgvector/pgvector-elixir)
@@ -392,8 +405,10 @@ Haskell | [pgvector-haskell](https://github.com/pgvector/pgvector-haskell)
Java, Kotlin, Groovy, Scala | [pgvector-java](https://github.com/pgvector/pgvector-java)
JavaScript, TypeScript | [pgvector-node](https://github.com/pgvector/pgvector-node)
Julia | [pgvector-julia](https://github.com/pgvector/pgvector-julia)
Lisp | [pgvector-lisp](https://github.com/pgvector/pgvector-lisp)
Lua | [pgvector-lua](https://github.com/pgvector/pgvector-lua)
Nim | [pgvector-nim](https://github.com/pgvector/pgvector-nim)
OCaml | [pgvector-ocaml](https://github.com/pgvector/pgvector-ocaml)
Perl | [pgvector-perl](https://github.com/pgvector/pgvector-perl)
PHP | [pgvector-php](https://github.com/pgvector/pgvector-php)
Python | [pgvector-python](https://github.com/pgvector/pgvector-python)
@@ -466,6 +481,14 @@ and query with:
SELECT * FROM items ORDER BY embedding::vector(3) <-> '[3,1,2]' LIMIT 5;
```
#### Do indexes need to fit into memory?
No, but like other index types, youll likely see better performance if they do. You can get the size of an index with:
```sql
SELECT pg_size_pretty(pg_relation_size('index_name'));
```
## Troubleshooting
#### Why isnt a query using an index?
@@ -565,10 +588,10 @@ If compilation fails with `fatal error: postgres.h: No such file or directory`,
For Ubuntu and Debian, use:
```sh
sudo apt install postgresql-server-dev-15
sudo apt install postgresql-server-dev-16
```
Note: Replace `15` with your Postgres server version
Note: Replace `16` with your Postgres server version
### Windows
@@ -583,7 +606,7 @@ Note: The exact path will vary depending on your Visual Studio version and editi
Then use `nmake` to build:
```cmd
set "PGROOT=C:\Program Files\PostgreSQL\15"
set "PGROOT=C:\Program Files\PostgreSQL\16"
git clone --branch v0.5.1 https://github.com/pgvector/pgvector.git
cd pgvector
nmake /F Makefile.win
@@ -633,22 +656,22 @@ pgxn install vector
Debian and Ubuntu packages are available from the [PostgreSQL APT Repository](https://wiki.postgresql.org/wiki/Apt). Follow the [setup instructions](https://wiki.postgresql.org/wiki/Apt#Quickstart) and run:
```sh
sudo apt install postgresql-15-pgvector
sudo apt install postgresql-16-pgvector
```
Note: Replace `15` with your Postgres server version
Note: Replace `16` with your Postgres server version
### Yum
RPM packages are available from the [PostgreSQL Yum Repository](https://yum.postgresql.org/). Follow the [setup instructions](https://www.postgresql.org/download/linux/redhat/) for your distribution and run:
```sh
sudo yum install pgvector_15
sudo yum install pgvector_16
# or
sudo dnf install pgvector_15
sudo dnf install pgvector_16
```
Note: Replace `15` with your Postgres server version
Note: Replace `16` with your Postgres server version
### conda-forge
@@ -718,7 +741,6 @@ Thanks to:
- [k-means++: The Advantage of Careful Seeding](https://theory.stanford.edu/~sergei/papers/kMeansPP-soda.pdf)
- [Concept Decompositions for Large Sparse Text Data using Clustering](https://www.cs.utexas.edu/users/inderjit/public_papers/concept_mlj.pdf)
- [Efficient and Robust Approximate Nearest Neighbor Search using Hierarchical Navigable Small World Graphs](https://arxiv.org/ftp/arxiv/papers/1603/1603.09320.pdf)
- [HQANN: Efficient and Robust Similarity Search for Hybrid Queries with Structured and Unstructured Constraints](https://arxiv.org/pdf/2207.07940.pdf)
## History

View File

@@ -1,18 +0,0 @@
-- complain if script is sourced in psql, rather than via CREATE EXTENSION
\echo Use "ALTER EXTENSION vector UPDATE TO '0.6.0'" to load this file. \quit
CREATE FUNCTION hnsw_attribute_distance(integer, integer) RETURNS float8
AS 'MODULE_PATHNAME', 'hnsw_int4_attribute_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
CREATE FUNCTION hnsw_attribute_distance(bigint, bigint) RETURNS float8
AS 'MODULE_PATHNAME', 'hnsw_int8_attribute_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
CREATE OPERATOR CLASS vector_integer_ops
DEFAULT FOR TYPE integer USING hnsw AS
OPERATOR 2 = (integer, integer),
FUNCTION 3 hnsw_attribute_distance(integer, integer);
CREATE OPERATOR CLASS vector_bigint_ops
DEFAULT FOR TYPE bigint USING hnsw AS
OPERATOR 2 = (bigint, bigint),
FUNCTION 3 hnsw_attribute_distance(bigint, bigint);

View File

@@ -290,21 +290,3 @@ CREATE OPERATOR CLASS vector_cosine_ops
OPERATOR 1 <=> (vector, vector) FOR ORDER BY float_ops,
FUNCTION 1 vector_negative_inner_product(vector, vector),
FUNCTION 2 vector_norm(vector);
-- hnsw attributes
CREATE FUNCTION hnsw_attribute_distance(integer, integer) RETURNS float8
AS 'MODULE_PATHNAME', 'hnsw_int4_attribute_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
CREATE FUNCTION hnsw_attribute_distance(bigint, bigint) RETURNS float8
AS 'MODULE_PATHNAME', 'hnsw_int8_attribute_distance' LANGUAGE C IMMUTABLE STRICT PARALLEL SAFE;
CREATE OPERATOR CLASS vector_integer_ops
DEFAULT FOR TYPE integer USING hnsw AS
OPERATOR 2 = (integer, integer),
FUNCTION 3 hnsw_attribute_distance(integer, integer);
CREATE OPERATOR CLASS vector_bigint_ops
DEFAULT FOR TYPE bigint USING hnsw AS
OPERATOR 2 = (bigint, bigint),
FUNCTION 3 hnsw_attribute_distance(bigint, bigint);

View File

@@ -14,14 +14,45 @@
#endif
int hnsw_ef_search;
int hnsw_lock_tranche_id;
static relopt_kind hnsw_relopt_kind;
/*
* Assign a tranche ID for our LWLocks. This only needs to be done by one
* backend, as the tranche ID is remembered in shared memory.
*
* This shared memory area is very small, so we just allocate it from the
* "slop" that PostgreSQL reserves for small allocations like this. If
* this grows bigger, we should use a shmem_request_hook and
* RequestAddinShmemSpace() to pre-reserve space for this.
*/
static void
HnswInitLockTranche(void)
{
int *tranche_ids;
bool found;
LWLockAcquire(AddinShmemInitLock, LW_EXCLUSIVE);
tranche_ids = ShmemInitStruct("hnsw LWLock ids",
sizeof(int) * 1,
&found);
if (!found)
tranche_ids[0] = LWLockNewTrancheId();
hnsw_lock_tranche_id = tranche_ids[0];
LWLockRelease(AddinShmemInitLock);
/* Per-backend registration of the tranche ID */
LWLockRegisterTranche(hnsw_lock_tranche_id, "HnswBuild");
}
/*
* Initialize index options and variables
*/
void
HnswInit(void)
{
HnswInitLockTranche();
hnsw_relopt_kind = add_reloption_kind();
add_int_reloption(hnsw_relopt_kind, "m", "Max number of connections",
HNSW_DEFAULT_M, HNSW_MIN_M, HNSW_MAX_M
@@ -167,7 +198,7 @@ hnswhandler(PG_FUNCTION_ARGS)
IndexAmRoutine *amroutine = makeNode(IndexAmRoutine);
amroutine->amstrategies = 0;
amroutine->amsupport = 3;
amroutine->amsupport = 2;
#if PG_VERSION_NUM >= 130000
amroutine->amoptsprocnum = 0;
#endif
@@ -175,7 +206,7 @@ hnswhandler(PG_FUNCTION_ARGS)
amroutine->amcanorderbyop = true;
amroutine->amcanbackward = false; /* can change direction mid-scan */
amroutine->amcanunique = false;
amroutine->amcanmulticol = true;
amroutine->amcanmulticol = false;
amroutine->amoptionalkey = true;
amroutine->amsearcharray = false;
amroutine->amsearchnulls = false;
@@ -222,31 +253,3 @@ hnswhandler(PG_FUNCTION_ARGS)
PG_RETURN_POINTER(amroutine);
}
/*
* Get the distance between two int4 attributes
*/
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnsw_int4_attribute_distance);
Datum
hnsw_int4_attribute_distance(PG_FUNCTION_ARGS)
{
int32 a = PG_GETARG_INT32(0);
int32 b = PG_GETARG_INT32(1);
double distance = ((double) a) - ((double) b);
PG_RETURN_FLOAT8(distance);
}
/*
* Get the distance between two int8 attributes
*/
PGDLLEXPORT PG_FUNCTION_INFO_V1(hnsw_int8_attribute_distance);
Datum
hnsw_int8_attribute_distance(PG_FUNCTION_ARGS)
{
int64 a = PG_GETARG_INT64(0);
int64 b = PG_GETARG_INT64(1);
double distance = ((double) a) - ((double) b);
PG_RETURN_FLOAT8(distance);
}

View File

@@ -4,9 +4,11 @@
#include "postgres.h"
#include "access/generic_xlog.h"
#include "access/parallel.h"
#include "access/reloptions.h"
#include "nodes/execnodes.h"
#include "port.h" /* for random() */
#include "utils/relptr.h"
#include "utils/sampling.h"
#include "vector.h"
@@ -14,12 +16,15 @@
#error "Requires PostgreSQL 11+"
#endif
#if PG_VERSION_NUM < 120000
#include "access/relscan.h"
#endif
#define HNSW_MAX_DIM 2000
/* Support functions */
#define HNSW_DISTANCE_PROC 1
#define HNSW_NORM_PROC 2
#define HNSW_ATTRIBUTE_DISTANCE_PROC 3
#define HNSW_VERSION 1
#define HNSW_MAGIC_NUMBER 0xA953A953
@@ -68,8 +73,10 @@
#if PG_VERSION_NUM >= 150000
#define RandomDouble() pg_prng_double(&pg_global_prng_state)
#define SeedRandom(seed) pg_prng_seed(&pg_global_prng_state, seed)
#else
#define RandomDouble() (((double) random()) / MAX_RANDOM_VALUE)
#define SeedRandom(seed) srandom(seed)
#endif
#if PG_VERSION_NUM < 130000
@@ -87,32 +94,64 @@
#define HnswGetMl(m) (1 / log(m))
/* Ensure fits on page and in uint8 */
#define HnswGetMaxLevel(m) Min(((BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData)) - offsetof(HnswNeighborTupleData, indextids) - sizeof(ItemIdData)) / (sizeof(ItemPointerData)) / m) - 2, 255)
#define HnswGetMaxLevel(m) Min(((BLCKSZ - MAXALIGN(SizeOfPageHeaderData) - MAXALIGN(sizeof(HnswPageOpaqueData)) - offsetof(HnswNeighborTupleData, indextids) - sizeof(ItemIdData)) / (sizeof(ItemPointerData)) / (m)) - 2, 255)
#define HnswGetValue(base, element) PointerGetDatum(HnswPtrAccess(base, (element)->value))
#if PG_VERSION_NUM < 140005
#define relptr_offset(rp) ((rp).relptr_off - 1)
#endif
/* Pointer macros */
#define HnswPtrAccess(base, hp) ((base) == NULL ? (hp).ptr : relptr_access(base, (hp).relptr))
#define HnswPtrStore(base, hp, value) ((base) == NULL ? (void) ((hp).ptr = (value)) : (void) relptr_store(base, (hp).relptr, value))
#define HnswPtrIsNull(base, hp) ((base) == NULL ? (hp).ptr == NULL : relptr_is_null((hp).relptr))
#define HnswPtrEqual(base, hp1, hp2) ((base) == NULL ? (hp1).ptr == (hp2).ptr : relptr_offset((hp1).relptr) == relptr_offset((hp2).relptr))
/* For code paths dedicated to each type */
#define HnswPtrPointer(hp) (hp).ptr
#define HnswPtrOffset(hp) relptr_offset((hp).relptr)
/* Variables */
extern int hnsw_ef_search;
extern int hnsw_lock_tranche_id;
typedef struct HnswElementData HnswElementData;
typedef struct HnswNeighborArray HnswNeighborArray;
#define HnswPtrDeclare(type, relptrtype, ptrtype) \
relptr_declare(type, relptrtype); \
typedef union { type *ptr; relptrtype relptr; } ptrtype;
/* Pointers that can be absolute or relative */
/* Use char for DatumPtr so works with Pointer */
HnswPtrDeclare(HnswElementData, HnswElementRelptr, HnswElementPtr);
HnswPtrDeclare(HnswNeighborArray, HnswNeighborArrayRelptr, HnswNeighborArrayPtr);
HnswPtrDeclare(HnswNeighborArrayPtr, HnswNeighborsRelptr, HnswNeighborsPtr);
HnswPtrDeclare(char, DatumRelptr, DatumPtr);
typedef struct HnswElementData
{
List *heaptids;
HnswElementPtr next;
ItemPointerData heaptids[HNSW_HEAPTIDS];
uint8 heaptidsLength;
uint8 level;
uint8 deleted;
HnswNeighborArray *neighbors;
uint32 hash;
HnswNeighborsPtr neighbors;
BlockNumber blkno;
OffsetNumber offno;
OffsetNumber neighborOffno;
BlockNumber neighborPage;
Datum value;
IndexTuple itup;
DatumPtr value;
LWLock lock;
} HnswElementData;
typedef HnswElementData * HnswElement;
typedef struct HnswCandidate
{
HnswElement element;
HnswElementPtr element;
float distance;
bool closer;
} HnswCandidate;
@@ -121,7 +160,7 @@ typedef struct HnswNeighborArray
{
int length;
bool closerSet;
HnswCandidate *items;
HnswCandidate items[FLEXIBLE_ARRAY_MEMBER];
} HnswNeighborArray;
typedef struct HnswPairingHeapNode
@@ -138,6 +177,70 @@ typedef struct HnswOptions
int efConstruction; /* size of dynamic candidate list */
} HnswOptions;
typedef struct HnswGraph
{
/* Graph state */
slock_t lock;
HnswElementPtr head;
double indtuples;
/* Entry state */
LWLock entryLock;
HnswElementPtr entryPoint;
/* Allocations state */
LWLock allocatorLock;
long memoryUsed;
long memoryTotal;
/* Flushed state */
LWLock flushLock;
bool flushed;
} HnswGraph;
typedef struct HnswShared
{
/* Immutable state */
Oid heaprelid;
Oid indexrelid;
bool isconcurrent;
/* Worker progress */
ConditionVariable workersdonecv;
/* Mutex for mutable state */
slock_t mutex;
/* Mutable state */
int nparticipantsdone;
double reltuples;
HnswGraph graphData;
#if PG_VERSION_NUM < 120000
ParallelHeapScanDescData heapdesc; /* must come last */
#endif
} HnswShared;
#if PG_VERSION_NUM >= 120000
#define ParallelTableScanFromHnswShared(shared) \
(ParallelTableScanDesc) ((char *) (shared) + BUFFERALIGN(sizeof(HnswShared)))
#endif
typedef struct HnswLeader
{
ParallelContext *pcxt;
int nparticipanttuplesorts;
HnswShared *hnswshared;
Snapshot snapshot;
char *hnswarea;
} HnswLeader;
typedef struct HnswAllocator
{
void *(*alloc) (Size size, void *state);
void *state;
} HnswAllocator;
typedef struct HnswBuildState
{
/* Info */
@@ -156,22 +259,26 @@ typedef struct HnswBuildState
double reltuples;
/* Support functions */
FmgrInfo **procinfos;
FmgrInfo *procinfo;
FmgrInfo *normprocinfo;
Oid *collations;
Oid collation;
/* Variables */
List *elements;
HnswElement entryPoint;
HnswGraph graphData;
HnswGraph *graph;
double ml;
int maxLevel;
long memoryLeft;
bool flushed;
bool useIndexTuple;
Vector *normvec;
/* Memory */
MemoryContext graphCtx;
MemoryContext tmpCtx;
HnswAllocator allocator;
/* Parallel builds */
HnswLeader *hnswleader;
HnswShared *hnswshared;
char *hnswarea;
} HnswBuildState;
typedef struct HnswMetaPageData
@@ -229,9 +336,9 @@ typedef struct HnswScanOpaqueData
MemoryContext tmpCtx;
/* Support functions */
FmgrInfo **procinfos;
FmgrInfo *procinfo;
FmgrInfo *normprocinfo;
Oid *collations;
Oid collation;
} HnswScanOpaqueData;
typedef HnswScanOpaqueData * HnswScanOpaque;
@@ -249,11 +356,11 @@ typedef struct HnswVacuumState
int efConstruction;
/* Support functions */
FmgrInfo **procinfos;
Oid *collations;
FmgrInfo *procinfo;
Oid collation;
/* Variables */
HTAB *deleted;
struct tidhash_hash *deleted;
BufferAccessStrategy bas;
HnswNeighborTuple ntup;
HnswElementData highestPoint;
@@ -267,32 +374,29 @@ int HnswGetM(Relation index);
int HnswGetEfConstruction(Relation index);
FmgrInfo *HnswOptionalProcInfo(Relation index, uint16 procnum);
bool HnswNormValue(FmgrInfo *procinfo, Oid collation, Datum *value, Vector * result);
void HnswCommitBuffer(Buffer buf, GenericXLogState *state);
Buffer HnswNewBuffer(Relation index, ForkNumber forkNum);
void HnswInitPage(Buffer buf, Page page);
void HnswInitRegisterPage(Relation index, Buffer *buf, Page *page, GenericXLogState **state);
void HnswInit(void);
List *HnswSearchLayer(Datum q, IndexTuple qtup, ScanKeyData *keyData, List *ep, int ef, int lc, Relation index, FmgrInfo **procinfos, Oid *collations, int m, bool loadVec, HnswElement skipElement, bool inMemory);
List *HnswSearchLayer(char *base, Datum q, List *ep, int ef, int lc, Relation index, FmgrInfo *procinfo, Oid collation, int m, bool inserting, HnswElement skipElement);
HnswElement HnswGetEntryPoint(Relation index);
void HnswGetMetaPageInfo(Relation index, int *m, HnswElement * entryPoint);
HnswElement HnswInitElement(ItemPointer tid, int m, double ml, int maxLevel);
void HnswFreeElement(HnswElement element);
void *HnswAlloc(HnswAllocator * allocator, Size size);
HnswElement HnswInitElement(char *base, ItemPointer tid, int m, double ml, int maxLevel, HnswAllocator * alloc);
HnswElement HnswInitElementFromBlock(BlockNumber blkno, OffsetNumber offno);
void HnswInsertElement(HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo **procinfos, Oid *collations, int m, int efConstruction, bool existing, bool inMemory);
HnswElement HnswFindDuplicate(HnswElement e, Relation index);
HnswCandidate *HnswEntryCandidate(HnswElement em, Datum q, IndexTuple qtup, ScanKeyData *keyData, Relation rel, FmgrInfo **procinfos, Oid *collations, bool loadVec, bool inMemory);
void HnswUpdateMetaPage(Relation index, int updateEntry, HnswElement entryPoint, BlockNumber insertPage, ForkNumber forkNum);
void HnswSetNeighborTuple(HnswNeighborTuple ntup, HnswElement e, int m);
void HnswFindElementNeighbors(char *base, HnswElement element, HnswElement entryPoint, Relation index, FmgrInfo *procinfo, Oid collation, int m, int efConstruction, bool existing);
HnswCandidate *HnswEntryCandidate(char *base, HnswElement em, Datum q, Relation rel, FmgrInfo *procinfo, Oid collation, bool loadVec);
void HnswUpdateMetaPage(Relation index, int updateEntry, HnswElement entryPoint, BlockNumber insertPage, ForkNumber forkNum, bool building);
void HnswSetNeighborTuple(char *base, HnswNeighborTuple ntup, HnswElement e, int m);
void HnswAddHeapTid(HnswElement element, ItemPointer heaptid);
void HnswInitNeighbors(HnswElement element, int m);
bool HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel);
void HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, HnswElement e, int m, bool checkExisting);
void HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHeaptids, bool loadVec, Relation index);
void HnswLoadElement(HnswElement element, float *distance, Datum *q, IndexTuple qtup, ScanKeyData *keyData, Relation index, FmgrInfo **procinfos, Oid *collations, bool loadVec);
void HnswSetElementTuple(HnswElementTuple etup, HnswElement element, bool useIndexTuple);
void HnswUpdateConnection(HnswElement element, HnswCandidate * hc, int m, int lc, int *updateIdx, Relation index, FmgrInfo **procinfos, Oid *collations, bool inMemory);
void HnswInitNeighbors(char *base, HnswElement element, int m, HnswAllocator * alloc);
bool HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel, bool building);
void HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building);
void HnswLoadElementFromTuple(HnswElement element, HnswElementTuple etup, bool loadHeaptids, bool loadVec);
void HnswLoadElement(HnswElement element, float *distance, Datum *q, Relation index, FmgrInfo *procinfo, Oid collation, bool loadVec);
void HnswSetElementTuple(char *base, HnswElementTuple etup, HnswElement element);
void HnswUpdateConnection(char *base, HnswElement element, HnswCandidate * hc, int m, int lc, int *updateIdx, Relation index, FmgrInfo *procinfo, Oid collation);
void HnswLoadNeighbors(HnswElement element, Relation index, int m);
void HnswElementSetData(HnswElement element, Relation index, Datum value, Datum *values, bool *isnull);
PGDLLEXPORT void HnswParallelBuildMain(dsm_segment *seg, shm_toc *toc);
/* Index access methods */
IndexBuildResult *hnswbuild(Relation heap, Relation index, IndexInfo *indexInfo);
@@ -309,6 +413,55 @@ IndexScanDesc hnswbeginscan(Relation index, int nkeys, int norderbys);
void hnswrescan(IndexScanDesc scan, ScanKey keys, int nkeys, ScanKey orderbys, int norderbys);
bool hnswgettuple(IndexScanDesc scan, ScanDirection dir);
void hnswendscan(IndexScanDesc scan);
FmgrInfo **HnswInitProcinfos(Relation index);
static inline HnswNeighborArray *
HnswGetNeighbors(char *base, HnswElement element, int lc)
{
HnswNeighborArrayPtr *neighborList = HnswPtrAccess(base, element->neighbors);
Assert(element->level >= lc);
return HnswPtrAccess(base, neighborList[lc]);
}
/* Hash tables */
typedef struct TidHashEntry
{
ItemPointerData tid;
char status;
} TidHashEntry;
#define SH_PREFIX tidhash
#define SH_ELEMENT_TYPE TidHashEntry
#define SH_KEY_TYPE ItemPointerData
#define SH_SCOPE extern
#define SH_DECLARE
#include "lib/simplehash.h"
typedef struct PointerHashEntry
{
uintptr_t ptr;
char status;
} PointerHashEntry;
#define SH_PREFIX pointerhash
#define SH_ELEMENT_TYPE PointerHashEntry
#define SH_KEY_TYPE uintptr_t
#define SH_SCOPE extern
#define SH_DECLARE
#include "lib/simplehash.h"
typedef struct OffsetHashEntry
{
Size offset;
char status;
} OffsetHashEntry;
#define SH_PREFIX offsethash
#define SH_ELEMENT_TYPE OffsetHashEntry
#define SH_KEY_TYPE Size
#define SH_SCOPE extern
#define SH_DECLARE
#include "lib/simplehash.h"
#endif

File diff suppressed because it is too large Load Diff

View File

@@ -5,6 +5,7 @@
#include "hnsw.h"
#include "storage/bufmgr.h"
#include "storage/lmgr.h"
#include "utils/datum.h"
#include "utils/memutils.h"
/*
@@ -92,7 +93,7 @@ HnswFreeOffset(Relation index, Buffer buf, Page page, HnswElement element, Size
* Add a new page
*/
static void
HnswInsertAppendPage(Relation index, Buffer *nbuf, Page *npage, GenericXLogState *state, Page page)
HnswInsertAppendPage(Relation index, Buffer *nbuf, Page *npage, GenericXLogState *state, Page page, bool building)
{
/* Add a new page */
LockRelationForExtension(index, ExclusiveLock);
@@ -100,7 +101,11 @@ HnswInsertAppendPage(Relation index, Buffer *nbuf, Page *npage, GenericXLogState
UnlockRelationForExtension(index, ExclusiveLock);
/* Init new page */
*npage = GenericXLogRegisterBuffer(state, *nbuf, GENERIC_XLOG_FULL_IMAGE);
if (building)
*npage = BufferGetPage(*nbuf);
else
*npage = GenericXLogRegisterBuffer(state, *nbuf, GENERIC_XLOG_FULL_IMAGE);
HnswInitPage(*nbuf, *npage);
/* Update previous buffer */
@@ -111,7 +116,7 @@ HnswInsertAppendPage(Relation index, Buffer *nbuf, Page *npage, GenericXLogState
* Add to element and neighbor pages
*/
static void
WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPage, BlockNumber *updatedInsertPage)
AddElementOnDisk(Relation index, HnswElement e, int m, BlockNumber insertPage, BlockNumber *updatedInsertPage, bool building)
{
Buffer buf;
Page page;
@@ -129,10 +134,10 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
OffsetNumber freeOffno = InvalidOffsetNumber;
OffsetNumber freeNeighborOffno = InvalidOffsetNumber;
BlockNumber newInsertPage = InvalidBlockNumber;
bool useIndexTuple = IndexRelationGetNumberOfAttributes(index) > 1;
char *base = NULL;
/* Calculate sizes */
etupSize = HNSW_ELEMENT_TUPLE_SIZE(useIndexTuple ? IndexTupleSize(e->itup) : VARSIZE_ANY(DatumGetPointer(e->value)));
etupSize = HNSW_ELEMENT_TUPLE_SIZE(VARSIZE_ANY(HnswPtrAccess(base, e->value)));
ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(e->level, m);
combinedSize = etupSize + ntupSize + sizeof(ItemIdData);
maxSize = HNSW_MAX_SIZE;
@@ -140,11 +145,11 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
/* Prepare element tuple */
etup = palloc0(etupSize);
HnswSetElementTuple(etup, e, useIndexTuple);
HnswSetElementTuple(base, etup, e);
/* Prepare neighbor tuple */
ntup = palloc0(ntupSize);
HnswSetNeighborTuple(ntup, e, m);
HnswSetNeighborTuple(base, ntup, e, m);
/* Find a page (or two if needed) to insert the tuples */
for (;;)
@@ -152,8 +157,16 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
buf = ReadBuffer(index, currentPage);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
if (building)
{
state = NULL;
page = BufferGetPage(buf);
}
else
{
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
}
/* Keep track of first page where element at level 0 can fit */
if (!BlockNumberIsValid(newInsertPage) && PageGetFreeSpace(page) >= minCombinedSize)
@@ -173,7 +186,12 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
if (HnswFreeOffset(index, buf, page, e, ntupSize, &nbuf, &npage, &freeOffno, &freeNeighborOffno, &newInsertPage))
{
if (nbuf != buf)
npage = GenericXLogRegisterBuffer(state, nbuf, 0);
{
if (building)
npage = BufferGetPage(nbuf);
else
npage = GenericXLogRegisterBuffer(state, nbuf, 0);
}
break;
}
@@ -182,7 +200,7 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
/* Skip if both tuples can fit on the same page */
if (combinedSize > maxSize && PageGetFreeSpace(page) >= etupSize && !BlockNumberIsValid(HnswPageGetOpaque(page)->nextblkno))
{
HnswInsertAppendPage(index, &nbuf, &npage, state, page);
HnswInsertAppendPage(index, &nbuf, &npage, state, page, building);
break;
}
@@ -191,7 +209,8 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
if (BlockNumberIsValid(currentPage))
{
/* Move to next page */
GenericXLogAbort(state);
if (!building)
GenericXLogAbort(state);
UnlockReleaseBuffer(buf);
}
else
@@ -199,22 +218,33 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
Buffer newbuf;
Page newpage;
HnswInsertAppendPage(index, &newbuf, &newpage, state, page);
HnswInsertAppendPage(index, &newbuf, &newpage, state, page, building);
/* Commit */
GenericXLogFinish(state);
if (building)
MarkBufferDirty(buf);
else
GenericXLogFinish(state);
/* Unlock previous buffer */
UnlockReleaseBuffer(buf);
/* Prepare new buffer */
state = GenericXLogStart(index);
buf = newbuf;
page = GenericXLogRegisterBuffer(state, buf, 0);
if (building)
{
state = NULL;
page = BufferGetPage(buf);
}
else
{
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
}
/* Create new page for neighbors if needed */
if (PageGetFreeSpace(page) < combinedSize)
HnswInsertAppendPage(index, &nbuf, &npage, state, page);
HnswInsertAppendPage(index, &nbuf, &npage, state, page, building);
else
{
nbuf = buf;
@@ -268,7 +298,14 @@ WriteNewElementPages(Relation index, HnswElement e, int m, BlockNumber insertPag
}
/* Commit */
GenericXLogFinish(state);
if (building)
{
MarkBufferDirty(buf);
if (nbuf != buf)
MarkBufferDirty(nbuf);
}
else
GenericXLogFinish(state);
UnlockReleaseBuffer(buf);
if (nbuf != buf)
UnlockReleaseBuffer(nbuf);
@@ -302,12 +339,14 @@ ConnectionExists(HnswElement e, HnswNeighborTuple ntup, int startIdx, int lm)
* Update neighbors
*/
void
HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, HnswElement e, int m, bool checkExisting)
HnswUpdateNeighborsOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement e, int m, bool checkExisting, bool building)
{
char *base = NULL;
for (int lc = e->level; lc >= 0; lc--)
{
int lm = HnswGetLayerM(m, lc);
HnswNeighborArray *neighbors = &e->neighbors[lc];
HnswNeighborArray *neighbors = HnswGetNeighbors(base, e, lc);
for (int i = 0; i < neighbors->length; i++)
{
@@ -320,11 +359,12 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, H
Size ntupSize;
int idx = -1;
int startIdx;
OffsetNumber offno = hc->element->neighborOffno;
HnswElement neighborElement = HnswPtrAccess(base, hc->element);
OffsetNumber offno = neighborElement->neighborOffno;
/* Get latest neighbors since they may have changed */
/* Do not lock yet since selecting neighbors can take time */
HnswLoadNeighbors(hc->element, index, m);
HnswLoadNeighbors(neighborElement, index, m);
/*
* Could improve performance for vacuuming by checking neighbors
@@ -334,17 +374,25 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, H
*/
/* Select neighbors */
HnswUpdateConnection(e, hc, lm, lc, &idx, index, procinfos, collations, false);
HnswUpdateConnection(NULL, e, hc, lm, lc, &idx, index, procinfo, collation);
/* New element was not selected as a neighbor */
if (idx == -1)
continue;
/* Register page */
buf = ReadBuffer(index, hc->element->neighborPage);
buf = ReadBuffer(index, neighborElement->neighborPage);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
if (building)
{
state = NULL;
page = BufferGetPage(buf);
}
else
{
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
}
/* Get tuple */
itemid = PageGetItemId(page, offno);
@@ -352,7 +400,7 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, H
ntupSize = ItemIdGetLength(itemid);
/* Calculate index for update */
startIdx = (hc->element->level - lc) * m;
startIdx = (neighborElement->level - lc) * m;
/* Check for existing connection */
if (checkExisting && ConnectionExists(e, ntup, startIdx, lm))
@@ -386,9 +434,12 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, H
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
/* Commit */
GenericXLogFinish(state);
if (building)
MarkBufferDirty(buf);
else
GenericXLogFinish(state);
}
else
else if (!building)
GenericXLogAbort(state);
UnlockReleaseBuffer(buf);
@@ -400,7 +451,7 @@ HnswUpdateNeighborPages(Relation index, FmgrInfo **procinfos, Oid *collations, H
* Add a heap TID to an existing element
*/
static bool
HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
AddDuplicateOnDisk(Relation index, HnswElement element, HnswElement dup, bool building)
{
Buffer buf;
Page page;
@@ -413,8 +464,16 @@ HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
/* Read page */
buf = ReadBuffer(index, dup->blkno);
LockBuffer(buf, BUFFER_LOCK_EXCLUSIVE);
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
if (building)
{
state = NULL;
page = BufferGetPage(buf);
}
else
{
state = GenericXLogStart(index);
page = GenericXLogRegisterBuffer(state, buf, 0);
}
/* Find space */
itemid = PageGetItemId(page, dup->offno);
@@ -429,82 +488,97 @@ HnswAddDuplicate(Relation index, HnswElement element, HnswElement dup)
/* Either being deleted or we lost our chance to another backend */
if (i == 0 || i == HNSW_HEAPTIDS)
{
GenericXLogAbort(state);
if (!building)
GenericXLogAbort(state);
UnlockReleaseBuffer(buf);
return false;
}
/* Add heap TID */
etup->heaptids[i] = *((ItemPointer) linitial(element->heaptids));
etup->heaptids[i] = element->heaptids[0];
/* Overwrite tuple */
if (!PageIndexTupleOverwrite(page, dup->offno, (Item) etup, etupSize))
elog(ERROR, "failed to add index item to \"%s\"", RelationGetRelationName(index));
/* Commit */
GenericXLogFinish(state);
if (building)
MarkBufferDirty(buf);
else
GenericXLogFinish(state);
UnlockReleaseBuffer(buf);
return true;
}
/*
* Write changes to disk
* Find duplicate element
*/
static bool
FindDuplicateOnDisk(Relation index, HnswElement element, bool building)
{
char *base = NULL;
HnswNeighborArray *neighbors = HnswGetNeighbors(base, element, 0);
Datum value = HnswGetValue(base, element);
for (int i = 0; i < neighbors->length; i++)
{
HnswCandidate *neighbor = &neighbors->items[i];
HnswElement neighborElement = HnswPtrAccess(base, neighbor->element);
Datum neighborValue = HnswGetValue(base, neighborElement);
/* Exit early since ordered by distance */
if (!datumIsEqual(value, neighborValue, false, -1))
return false;
if (AddDuplicateOnDisk(index, element, neighborElement, building))
return true;
}
return false;
}
/*
* Update graph on disk
*/
static void
WriteElement(Relation index, FmgrInfo **procinfos, Oid *collations, HnswElement element, int m, int efConstruction, HnswElement dup, HnswElement entryPoint)
UpdateGraphOnDisk(Relation index, FmgrInfo *procinfo, Oid collation, HnswElement element, int m, int efConstruction, HnswElement entryPoint, bool building)
{
BlockNumber newInsertPage = InvalidBlockNumber;
/* Try to add to existing page */
if (dup != NULL)
{
if (HnswAddDuplicate(index, element, dup))
return;
}
/* Look for duplicate */
if (FindDuplicateOnDisk(index, element, building))
return;
/* Write element and neighbor tuples */
WriteNewElementPages(index, element, m, GetInsertPage(index), &newInsertPage);
/* Add element */
AddElementOnDisk(index, element, m, GetInsertPage(index), &newInsertPage, building);
/* Update insert page if needed */
if (BlockNumberIsValid(newInsertPage))
HnswUpdateMetaPage(index, 0, NULL, newInsertPage, MAIN_FORKNUM);
HnswUpdateMetaPage(index, 0, NULL, newInsertPage, MAIN_FORKNUM, building);
/* Update neighbors */
HnswUpdateNeighborPages(index, procinfos, collations, element, m, false);
HnswUpdateNeighborsOnDisk(index, procinfo, collation, element, m, false, building);
/* Update metapage if needed */
/* Update entry point if needed */
if (entryPoint == NULL || element->level > entryPoint->level)
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_GREATER, element, InvalidBlockNumber, MAIN_FORKNUM);
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_GREATER, element, InvalidBlockNumber, MAIN_FORKNUM, building);
}
/*
* Insert a tuple into the index
*/
bool
HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel)
HnswInsertTupleOnDisk(Relation index, Datum value, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel, bool building)
{
Datum value;
FmgrInfo *normprocinfo;
HnswElement entryPoint;
HnswElement element;
int m;
int efConstruction = HnswGetEfConstruction(index);
FmgrInfo **procinfos = HnswInitProcinfos(index);
Oid *collations = index->rd_indcollation;
HnswElement dup;
FmgrInfo *procinfo = index_getprocinfo(index, 1, HNSW_DISTANCE_PROC);
Oid collation = index->rd_indcollation[0];
LOCKMODE lockmode = ShareLock;
/* Detoast once for all calls */
value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
/* Normalize if needed */
normprocinfo = HnswOptionalProcInfo(index, HNSW_NORM_PROC);
if (normprocinfo != NULL)
{
if (!HnswNormValue(normprocinfo, collations[0], &value, NULL))
return false;
}
char *base = NULL;
/*
* Get a shared lock. This allows vacuum to ensure no in-flight inserts
@@ -517,8 +591,8 @@ HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_ti
HnswGetMetaPageInfo(index, &m, &entryPoint);
/* Create an element */
element = HnswInitElement(heap_tid, m, HnswGetMl(m), HnswGetMaxLevel(m));
HnswElementSetData(element, index, value, values, isnull);
element = HnswInitElement(base, heap_tid, m, HnswGetMl(m), HnswGetMaxLevel(m), NULL);
HnswPtrStore(base, element->value, DatumGetPointer(value));
/* Prevent concurrent inserts when likely updating entry point */
if (entryPoint == NULL || element->level > entryPoint->level)
@@ -534,14 +608,11 @@ HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_ti
entryPoint = HnswGetEntryPoint(index);
}
/* Insert element in graph */
HnswInsertElement(element, entryPoint, index, procinfos, collations, m, efConstruction, false, false);
/* Find neighbors for element */
HnswFindElementNeighbors(base, element, entryPoint, index, procinfo, collation, m, efConstruction, false);
/* Look for duplicate */
dup = HnswFindDuplicate(element, index);
/* Write to disk */
WriteElement(index, procinfos, collations, element, m, efConstruction, dup, entryPoint);
/* Update graph on disk */
UpdateGraphOnDisk(index, procinfo, collation, element, m, efConstruction, entryPoint, building);
/* Release lock */
UnlockPage(index, HNSW_UPDATE_LOCK, lockmode);
@@ -549,6 +620,30 @@ HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_ti
return true;
}
/*
* Insert a tuple into the index
*/
static void
HnswInsertTuple(Relation index, Datum *values, bool *isnull, ItemPointer heap_tid, Relation heapRel)
{
Datum value;
FmgrInfo *normprocinfo;
Oid collation = index->rd_indcollation[0];
/* Detoast once for all calls */
value = PointerGetDatum(PG_DETOAST_DATUM(values[0]));
/* Normalize if needed */
normprocinfo = HnswOptionalProcInfo(index, HNSW_NORM_PROC);
if (normprocinfo != NULL)
{
if (!HnswNormValue(normprocinfo, collation, &value, NULL))
return;
}
HnswInsertTupleOnDisk(index, value, values, isnull, heap_tid, heapRel, false);
}
/*
* Insert a tuple into the index
*/

View File

@@ -15,13 +15,13 @@ GetScanItems(IndexScanDesc scan, Datum q)
{
HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
Relation index = scan->indexRelation;
FmgrInfo **procinfos = so->procinfos;
Oid *collations = so->collations;
FmgrInfo *procinfo = so->procinfo;
Oid collation = so->collation;
List *ep;
List *w;
int m;
HnswElement entryPoint;
ScanKeyData *keyData = scan->keyData;
char *base = NULL;
/* Get m and entry point */
HnswGetMetaPageInfo(index, &m, &entryPoint);
@@ -29,15 +29,15 @@ GetScanItems(IndexScanDesc scan, Datum q)
if (entryPoint == NULL)
return NIL;
ep = list_make1(HnswEntryCandidate(entryPoint, q, NULL, keyData, index, procinfos, collations, false, false));
ep = list_make1(HnswEntryCandidate(base, entryPoint, q, index, procinfo, collation, false));
for (int lc = entryPoint->level; lc >= 1; lc--)
{
w = HnswSearchLayer(q, NULL, keyData, ep, 1, lc, index, procinfos, collations, m, false, NULL, false);
w = HnswSearchLayer(base, q, ep, 1, lc, index, procinfo, collation, m, false, NULL);
ep = w;
}
return HnswSearchLayer(q, NULL, keyData, ep, hnsw_ef_search, 0, index, procinfos, collations, m, false, NULL, false);
return HnswSearchLayer(base, q, ep, hnsw_ef_search, 0, index, procinfo, collation, m, false, NULL);
}
/*
@@ -84,7 +84,7 @@ GetScanValue(IndexScanDesc scan)
/* Fine if normalization fails */
if (so->normprocinfo != NULL)
HnswNormValue(so->normprocinfo, so->collations[0], &value, NULL);
HnswNormValue(so->normprocinfo, so->collation, &value, NULL);
}
return value;
@@ -108,9 +108,9 @@ hnswbeginscan(Relation index, int nkeys, int norderbys)
ALLOCSET_DEFAULT_SIZES);
/* Set support functions */
so->procinfos = HnswInitProcinfos(index);
so->procinfo = index_getprocinfo(index, 1, HNSW_DISTANCE_PROC);
so->normprocinfo = HnswOptionalProcInfo(index, HNSW_NORM_PROC);
so->collations = index->rd_indcollation;
so->collation = index->rd_indcollation[0];
scan->opaque = so;
@@ -185,19 +185,19 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
while (list_length(so->w) > 0)
{
char *base = NULL;
HnswCandidate *hc = llast(so->w);
HnswElement element = HnswPtrAccess(base, hc->element);
ItemPointer heaptid;
/* Move to next element if no valid heap TIDs */
if (list_length(hc->element->heaptids) == 0)
if (element->heaptidsLength == 0)
{
so->w = list_delete_last(so->w);
continue;
}
heaptid = llast(hc->element->heaptids);
hc->element->heaptids = list_delete_last(hc->element->heaptids);
heaptid = &element->heaptids[--element->heaptidsLength];
MemoryContextSwitchTo(oldCtx);
@@ -207,9 +207,6 @@ hnswgettuple(IndexScanDesc scan, ScanDirection dir)
scan->xs_ctup.t_self = *heaptid;
#endif
/* TODO Check during scan */
scan->xs_recheck = scan->numberOfKeys > 0;
scan->xs_recheckorderby = false;
return true;
}
@@ -226,7 +223,6 @@ hnswendscan(IndexScanDesc scan)
{
HnswScanOpaque so = (HnswScanOpaque) scan->opaque;
pfree(so->procinfos);
MemoryContextDelete(so->tmpCtx);
pfree(so);

File diff suppressed because it is too large Load Diff

View File

@@ -12,12 +12,9 @@
* Check if deleted list contains an index TID
*/
static bool
DeletedContains(HTAB *deleted, ItemPointer indextid)
DeletedContains(tidhash_hash * deleted, ItemPointer indextid)
{
bool found;
hash_search(deleted, indextid, HASH_FIND, &found);
return found;
return tidhash_lookup(deleted, *indextid) != NULL;
}
/*
@@ -110,11 +107,13 @@ RemoveHeapTids(HnswVacuumState * vacuumstate)
if (!ItemPointerIsValid(&etup->heaptids[0]))
{
ItemPointerData ip;
bool found;
/* Add to deleted list */
ItemPointerSet(&ip, blkno, offno);
(void) hash_search(vacuumstate->deleted, &ip, HASH_ENTER, NULL);
tidhash_insert(vacuumstate->deleted, ip, &found);
Assert(!found);
}
else if (etup->level > highestLevel && !(entryPoint != NULL && blkno == entryPoint->blkno && offno == entryPoint->offno))
{
@@ -195,26 +194,27 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
GenericXLogState *state;
int m = vacuumstate->m;
int efConstruction = vacuumstate->efConstruction;
FmgrInfo **procinfos = vacuumstate->procinfos;
Oid *collations = vacuumstate->collations;
FmgrInfo *procinfo = vacuumstate->procinfo;
Oid collation = vacuumstate->collation;
BufferAccessStrategy bas = vacuumstate->bas;
HnswNeighborTuple ntup = vacuumstate->ntup;
Size ntupSize = HNSW_NEIGHBOR_TUPLE_SIZE(element->level, m);
char *base = NULL;
/* Skip if element is entry point */
if (entryPoint != NULL && element->blkno == entryPoint->blkno && element->offno == entryPoint->offno)
return;
/* Init fields */
HnswInitNeighbors(element, m);
element->heaptids = NIL;
HnswInitNeighbors(base, element, m, NULL);
element->heaptidsLength = 0;
/* Add element to graph, skipping itself */
HnswInsertElement(element, entryPoint, index, procinfos, collations, m, efConstruction, true, false);
/* Find neighbors for element, skipping itself */
HnswFindElementNeighbors(base, element, entryPoint, index, procinfo, collation, m, efConstruction, true);
/* Update neighbor tuple */
/* Do this before getting page to minimize locking */
HnswSetNeighborTuple(ntup, element, m);
HnswSetNeighborTuple(base, ntup, element, m);
/* Get neighbor page */
buf = ReadBufferExtended(index, MAIN_FORKNUM, element->neighborPage, RBM_NORMAL, bas);
@@ -231,7 +231,7 @@ RepairGraphElement(HnswVacuumState * vacuumstate, HnswElement element, HnswEleme
UnlockReleaseBuffer(buf);
/* Update neighbors */
HnswUpdateNeighborPages(index, procinfos, collations, element, m, true);
HnswUpdateNeighborsOnDisk(index, procinfo, collation, element, m, true, false);
}
/*
@@ -258,7 +258,7 @@ RepairGraphEntryPoint(HnswVacuumState * vacuumstate)
LockPage(index, HNSW_UPDATE_LOCK, ShareLock);
/* Load element */
HnswLoadElement(highestPoint, NULL, NULL, NULL, NULL, index, vacuumstate->procinfos, vacuumstate->collations, true);
HnswLoadElement(highestPoint, NULL, NULL, index, vacuumstate->procinfo, vacuumstate->collation, true);
/* Repair if needed */
if (NeedsUpdated(vacuumstate, highestPoint))
@@ -287,7 +287,7 @@ RepairGraphEntryPoint(HnswVacuumState * vacuumstate)
* point is outdated and empty, the entry point will be empty
* until an element is repaired.
*/
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_ALWAYS, highestPoint, InvalidBlockNumber, MAIN_FORKNUM);
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_ALWAYS, highestPoint, InvalidBlockNumber, MAIN_FORKNUM, false);
}
else
{
@@ -296,13 +296,13 @@ RepairGraphEntryPoint(HnswVacuumState * vacuumstate)
* is outdated, this can remove connections at higher levels in
* the graph until they are repaired, but this should be fine.
*/
HnswLoadElement(entryPoint, NULL, NULL, NULL, NULL, index, vacuumstate->procinfos, vacuumstate->collations, true);
HnswLoadElement(entryPoint, NULL, NULL, index, vacuumstate->procinfo, vacuumstate->collation, true);
if (NeedsUpdated(vacuumstate, entryPoint))
{
/* Reset neighbors from previous update */
if (highestPoint != NULL)
highestPoint->neighbors = NULL;
HnswPtrStore((char *) NULL, highestPoint->neighbors, (HnswNeighborArrayPtr *) NULL);
RepairGraphElement(vacuumstate, entryPoint, highestPoint);
}
@@ -372,7 +372,7 @@ RepairGraph(HnswVacuumState * vacuumstate)
/* Create an element */
element = HnswInitElementFromBlock(blkno, offno);
HnswLoadElementFromTuple(element, etup, false, true, index);
HnswLoadElementFromTuple(element, etup, false, true);
elements = lappend(elements, element);
}
@@ -420,7 +420,7 @@ RepairGraph(HnswVacuumState * vacuumstate)
* was replaced and highest point was outdated.
*/
if (entryPoint == NULL || element->level > entryPoint->level)
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_GREATER, element, InvalidBlockNumber, MAIN_FORKNUM);
HnswUpdateMetaPage(index, HNSW_UPDATE_ENTRY_GREATER, element, InvalidBlockNumber, MAIN_FORKNUM, false);
/* Release lock */
UnlockPage(index, HNSW_UPDATE_LOCK, lockmode);
@@ -442,7 +442,6 @@ MarkDeleted(HnswVacuumState * vacuumstate)
BlockNumber insertPage = InvalidBlockNumber;
Relation index = vacuumstate->index;
BufferAccessStrategy bas = vacuumstate->bas;
bool useIndexTuple = IndexRelationGetNumberOfAttributes(index);
/*
* Wait for index scans to complete. Scans before this point may contain
@@ -531,18 +530,7 @@ MarkDeleted(HnswVacuumState * vacuumstate)
/* Overwrite element */
etup->deleted = 1;
if (useIndexTuple)
{
IndexTuple itup = (IndexTuple) &etup->data;
MemSet(itup, 0, IndexTupleSize(itup));
}
else
{
Vector *vec = (Vector *) (&etup->data);
MemSet(vec, 0, VARSIZE_ANY(vec));
}
MemSet(&etup->data, 0, VARSIZE_ANY(&etup->data));
/* Overwrite neighbors */
for (int i = 0; i < ntup->count; i++)
@@ -577,7 +565,7 @@ MarkDeleted(HnswVacuumState * vacuumstate)
}
/* Update insert page last, after everything has been marked as deleted */
HnswUpdateMetaPage(index, 0, NULL, insertPage, MAIN_FORKNUM);
HnswUpdateMetaPage(index, 0, NULL, insertPage, MAIN_FORKNUM, false);
}
/*
@@ -587,7 +575,6 @@ static void
InitVacuumState(HnswVacuumState * vacuumstate, IndexVacuumInfo *info, IndexBulkDeleteResult *stats, IndexBulkDeleteCallback callback, void *callback_state)
{
Relation index = info->index;
HASHCTL hash_ctl;
if (stats == NULL)
stats = (IndexBulkDeleteResult *) palloc0(sizeof(IndexBulkDeleteResult));
@@ -598,8 +585,8 @@ InitVacuumState(HnswVacuumState * vacuumstate, IndexVacuumInfo *info, IndexBulkD
vacuumstate->callback_state = callback_state;
vacuumstate->efConstruction = HnswGetEfConstruction(index);
vacuumstate->bas = GetAccessStrategy(BAS_BULKREAD);
vacuumstate->procinfos = HnswInitProcinfos(index);
vacuumstate->collations = index->rd_indcollation;
vacuumstate->procinfo = index_getprocinfo(index, 1, HNSW_DISTANCE_PROC);
vacuumstate->collation = index->rd_indcollation[0];
vacuumstate->ntup = palloc0(BLCKSZ);
vacuumstate->tmpCtx = AllocSetContextCreate(CurrentMemoryContext,
"Hnsw vacuum temporary context",
@@ -609,10 +596,7 @@ InitVacuumState(HnswVacuumState * vacuumstate, IndexVacuumInfo *info, IndexBulkD
HnswGetMetaPageInfo(index, &vacuumstate->m, NULL);
/* Create hash table */
hash_ctl.keysize = sizeof(ItemPointerData);
hash_ctl.entrysize = sizeof(ItemPointerData);
hash_ctl.hcxt = CurrentMemoryContext;
vacuumstate->deleted = hash_create("hnswbulkdelete indextids", 256, &hash_ctl, HASH_ELEM | HASH_BLOBS | HASH_CONTEXT);
vacuumstate->deleted = tidhash_create(CurrentMemoryContext, 256, NULL);
}
/*
@@ -621,9 +605,8 @@ InitVacuumState(HnswVacuumState * vacuumstate, IndexVacuumInfo *info, IndexBulkD
static void
FreeVacuumState(HnswVacuumState * vacuumstate)
{
hash_destroy(vacuumstate->deleted);
tidhash_destroy(vacuumstate->deleted);
FreeAccessStrategy(vacuumstate->bas);
pfree(vacuumstate->procinfos);
pfree(vacuumstate->ntup);
MemoryContextDelete(vacuumstate->tmpCtx);
}

View File

@@ -543,10 +543,10 @@ CreateListPages(Relation index, VectorArray centers, int dimensions,
pfree(list);
}
#ifdef IVFFLAT_KMEANS_DEBUG
/*
* Print k-means metrics
*/
#ifdef IVFFLAT_KMEANS_DEBUG
static void
PrintKmeansMetrics(IvfflatBuildState * buildstate)
{

View File

@@ -6,6 +6,10 @@
#include "ivfflat.h"
#include "miscadmin.h"
#ifdef IVFFLAT_MEMORY
#include "utils/memutils.h"
#endif
/*
* Initialize with kmeans++
*
@@ -151,6 +155,23 @@ QuickCenters(Relation index, VectorArray samples, VectorArray centers)
}
}
#ifdef IVFFLAT_MEMORY
/*
* Show memory usage
*/
static void
ShowMemoryUsage(Size estimatedSize)
{
#if PG_VERSION_NUM >= 130000
elog(INFO, "total memory: %zu MB",
MemoryContextMemAllocated(CurrentMemoryContext, true) / (1024 * 1024));
#else
MemoryContextStats(CurrentMemoryContext);
#endif
elog(INFO, "estimated memory: %zu MB", estimatedSize / (1024 * 1024));
}
#endif
/*
* Use Elkan for performance. This requires distance function to satisfy triangle inequality.
*
@@ -231,6 +252,10 @@ ElkanKmeans(Relation index, VectorArray samples, VectorArray centers)
vec->dim = dimensions;
}
#ifdef IVFFLAT_MEMORY
ShowMemoryUsage(totalSize);
#endif
/* Pick initial centers */
InitCenters(index, samples, centers, lowerBound);

View File

@@ -724,7 +724,7 @@ vector_spherical_distance(PG_FUNCTION_ARGS)
}
/*
* Get the L1 distance between vectors
* Get the L1 distance between two vectors
*/
PGDLLEXPORT PG_FUNCTION_INFO_V1(l1_distance);
Datum
@@ -903,8 +903,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_lt);
Datum
vector_lt(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) < 0);
}
@@ -916,8 +916,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_le);
Datum
vector_le(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) <= 0);
}
@@ -929,8 +929,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_eq);
Datum
vector_eq(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) == 0);
}
@@ -942,8 +942,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ne);
Datum
vector_ne(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) != 0);
}
@@ -955,8 +955,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_ge);
Datum
vector_ge(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) >= 0);
}
@@ -968,8 +968,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_gt);
Datum
vector_gt(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_BOOL(vector_cmp_internal(a, b) > 0);
}
@@ -981,8 +981,8 @@ PGDLLEXPORT PG_FUNCTION_INFO_V1(vector_cmp);
Datum
vector_cmp(PG_FUNCTION_ARGS)
{
Vector *a = (Vector *) PG_GETARG_VECTOR_P(0);
Vector *b = (Vector *) PG_GETARG_VECTOR_P(1);
Vector *a = PG_GETARG_VECTOR_P(0);
Vector *b = PG_GETARG_VECTOR_P(1);
PG_RETURN_INT32(vector_cmp_internal(a, b));
}

View File

@@ -83,11 +83,46 @@ for my $i (0 .. $#operators)
push(@expected, $res);
}
# Add index
$node->safe_psql("postgres", "CREATE INDEX ON tst USING hnsw (v $opclass);");
# Build index serially
$node->safe_psql("postgres", qq(
SET max_parallel_maintenance_workers = 0;
CREATE INDEX idx ON tst USING hnsw (v $opclass);
));
# Test approximate results
my $min = $operator eq "<#>" ? 0.80 : 0.99;
test_recall($min, $operator);
$node->safe_psql("postgres", "DROP INDEX idx;");
# Build index in parallel in memory
my ($ret, $stdout, $stderr) = $node->psql("postgres", qq(
SET client_min_messages = DEBUG;
SET min_parallel_table_scan_size = 1;
CREATE INDEX idx ON tst USING hnsw (v $opclass);
));
is($ret, 0, $stderr);
like($stderr, qr/using \d+ parallel workers/);
# Test approximate results
test_recall($min, $operator);
$node->safe_psql("postgres", "DROP INDEX idx;");
# Build index in parallel on disk
# Set parallel_workers on table to use workers with low maintenance_work_mem
($ret, $stdout, $stderr) = $node->psql("postgres", qq(
ALTER TABLE tst SET (parallel_workers = 2);
SET client_min_messages = DEBUG;
SET maintenance_work_mem = '4MB';
CREATE INDEX idx ON tst USING hnsw (v $opclass);
ALTER TABLE tst RESET (parallel_workers);
));
is($ret, 0, $stderr);
like($stderr, qr/using \d+ parallel workers/);
like($stderr, qr/hnsw graph no longer fits into maintenance_work_mem/);
$node->safe_psql("postgres", "DROP INDEX idx;");
}
done_testing();

View File

@@ -1,107 +0,0 @@
use strict;
use warnings;
use PostgresNode;
use TestLib;
use Test::More;
my $node;
my @queries = ();
my @cs = ();
my @expected;
my $limit = 20;
my $dim = 3;
my $array_sql = join(",", ('random()') x $dim);
my $nc = 50;
sub test_recall
{
my ($min, $operator) = @_;
my $correct = 0;
my $total = 0;
my $explain = $node->safe_psql("postgres", qq(
SET enable_seqscan = off;
EXPLAIN ANALYZE SELECT i FROM tst WHERE c = $cs[0] ORDER BY v $operator '$queries[0]' LIMIT $limit;
));
like($explain, qr/Index Cond/);
for my $i (0 .. $#queries)
{
my $actual = $node->safe_psql("postgres", qq(
SET enable_seqscan = off;
SELECT i FROM tst WHERE c = $cs[$i] ORDER BY v $operator '$queries[$i]' LIMIT $limit;
));
my @actual_ids = split("\n", $actual);
my %actual_set = map { $_ => 1 } @actual_ids;
my @expected_ids = split("\n", $expected[$i]);
foreach (@expected_ids)
{
if (exists($actual_set{$_}))
{
$correct++;
}
$total++;
}
}
cmp_ok($correct / $total, ">=", $min, $operator);
}
# Initialize node
$node = get_new_node('node');
$node->init;
$node->start;
# Create table
$node->safe_psql("postgres", "CREATE EXTENSION vector;");
$node->safe_psql("postgres", "CREATE TABLE tst (i int4, v vector($dim), c int8);");
$node->safe_psql("postgres",
"INSERT INTO tst SELECT i, ARRAY[$array_sql], i % $nc FROM generate_series(1, 20000) i;"
);
# Generate queries
for (1 .. 20)
{
my @r = ();
for (1 .. $dim)
{
push(@r, rand());
}
push(@queries, "[" . join(",", @r) . "]");
push(@cs, int(rand() * $nc));
}
# Get exact results
@expected = ();
for my $i (0 .. $#queries)
{
my $res = $node->safe_psql("postgres", "SELECT i FROM tst WHERE c = $cs[$i] ORDER BY v <-> '$queries[$i]' LIMIT $limit;");
push(@expected, $res);
}
# Add index
$node->safe_psql("postgres", "CREATE INDEX ON tst USING hnsw (v vector_l2_ops, c);");
# Test recall
test_recall(0.99, '<->');
# Test vacuum
$node->safe_psql("postgres", "DELETE FROM tst WHERE c > 5;");
$node->safe_psql("postgres", "VACUUM tst;");
# Test columns
my ($ret, $stdout, $stderr) = $node->psql("postgres", "CREATE INDEX ON tst USING hnsw (c);");
like($stderr, qr/first column must be a vector/);
($ret, $stdout, $stderr) = $node->psql("postgres", "CREATE INDEX ON tst USING hnsw (c, v vector_l2_ops);");
like($stderr, qr/first column must be a vector/);
($ret, $stdout, $stderr) = $node->psql("postgres", "CREATE INDEX ON tst USING hnsw (v vector_l2_ops, c, c);");
like($stderr, qr/index cannot have more than two columns/);
($ret, $stdout, $stderr) = $node->psql("postgres", "CREATE INDEX ON tst USING hnsw (v vector_l2_ops, v vector_l2_ops);");
like($stderr, qr/column 2 cannot be a vector/);
done_testing();