blob: dd0541eab33e33651b11f97412fd89edbc7da3ce (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
|
---
slug: pg-trigger-tests
title: "Postgres Trigger Tests"
date: "2022-01-23T10:01:49-04:00"
---
We already know statement-level `INSERT` [triggers][tr] are faster than
row-level [triggers][tr] in [Postgres][].
Out of curiosity I decided to time row-level and statement-level
`INSERT` triggers with a variety of row counts.
Here are the results:
{{< figure
src="/files/posts/pg-trigger-tests/results.min.svg"
class=image
caption="Row Count vs Query Time"
>}}
The takeaway here is that if speed is a concern, then you should prefer
statement-level [triggers][tr] to row-level [triggers][tr].
The scripts used to create the test database, run the tests, and
generate the results are available in the [companion GitHub
repository][repo].
[docker]: https://docker.com/
"Docker container orchestrator."
[ruby]: https://ruby-lang.org/
"Ruby programming language."
[python]: https://python.org/
"Python programming language."
[matplotlib]: https://matplotlib.org/
"Matplotlib plotting library."
[tr]: https://www.postgresql.org/docs/current/sql-createtrigger.html
"Postgres CREATE TRIGGER statement."
[csv]: https://en.wikipedia.org/wiki/Comma-separated_values
"Comma-separated values"
[svg]: https://en.wikipedia.org/wiki/Scalable_Vector_Graphics
"Scalable Vector Graphics"
[postgres]: https://www.postgresql.org/
"PostgreSQL relational database."
[repo]: https://github.com/pablotron/pg-trigger-tests
"Companion GitHub repository."
|
