zihaozeng0021 opened a new issue, #2288:
URL: https://github.com/apache/age/issues/2288
### Summary
This report highlights significant performance discrepancies between
logically equivalent Cypher queries that should all return empty results.
Although every query correctly returns 0 rows, some take way much longer to
execute despite having the same semantic meaning.
### Configuration
AGE Version: `apache/age:latest` (pulled 2025-12-26)
Operating System: WSL2 – Ubuntu 24.04.1 LTS
Installation Method: Docker
---
### Steps to reproduce
1. Start AGE via Docker
```
docker run \
--name age \
-p 5455:5432 \
-e POSTGRES_USER=postgresUser \
-e POSTGRES_PASSWORD=postgresPW \
-e POSTGRES_DB=postgresDB \
-d \
apache/age:latest
```
2. Connect with psql
```
PGPASSWORD='postgresPW' psql -h localhost -p 5455 -U postgresUser -d
postgresDB
```
3. Run the following query
Set up:
```
\timing on
CREATE EXTENSION IF NOT EXISTS age;
LOAD 'age';
SET search_path = ag_catalog, "$user", public;
SELECT create_graph('test');
SELECT *
FROM cypher('test', $$
WITH range(1, 500000) AS ids
UNWIND ids AS id
CREATE (:Person {age: (id % 50) + 18, score: (id % 100)})
RETURN 1
$$) AS (ok agtype);
```
Positive case (finish immediately; All the negative cases below can be
simplified to this one):
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WHERE false
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 0.347 ms
```
Negative cases (take much longer):
Some of the buggy queries may share the same root cause; I reported all of
them anyway in case they turn out to be different issues.
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WHERE id(q) <> id(q)
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 565.713 ms
```
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WITH p, count(q) AS c
WHERE c < 0
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 307900.914 ms (05:07.901)
```
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WITH p, {n:q} AS m
WHERE m.n IS NULL
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 254.241 ms
```
The following queries are not taking much longer, I am not sure are they
really bugs:
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WHERE q.age = 1 AND q.age = 2
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 73.505 ms
```
```
postgresDB=# SELECT count(*)
FROM cypher('test', $$
MATCH (p:Person)
WITH p LIMIT 1000
MATCH (q:Person)
WHERE q.score > 50 AND q.score <= 50
RETURN p
$$) AS (p agtype);
count
-------
0
(1 row)
Time: 94.082 ms
```
---
### Expected behaviour
All queries should have comparable execution times, since each `WHERE
predicate` is a constant condition that cannot evaluate to true for any row.
### Actual behaviour
The positive case complete almost immediately. The negative cases take much
longer.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
