Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sokil/id-bench
Database identifier benchmarks
https://github.com/sokil/id-bench
autoincrement b-tree benchmarking mariadb mongodb mysql percona performance-analysis postgresql uuid uuid7
Last synced: 25 days ago
JSON representation
Database identifier benchmarks
- Host: GitHub
- URL: https://github.com/sokil/id-bench
- Owner: sokil
- License: mit
- Created: 2023-03-24T09:27:13.000Z (almost 2 years ago)
- Default Branch: main
- Last Pushed: 2023-07-28T08:43:52.000Z (over 1 year ago)
- Last Synced: 2024-10-28T11:04:09.788Z (2 months ago)
- Topics: autoincrement, b-tree, benchmarking, mariadb, mongodb, mysql, percona, performance-analysis, postgresql, uuid, uuid7
- Language: PHP
- Homepage:
- Size: 2.73 MB
- Stars: 2
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# id-bench
Primary key and secondary key insertion and selection benchmarking
## 1. UUID Versions
| Version | Description |
|---------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1 | The time-based version. |
| 2 | DCE Security version, with embedded POSIX UIDs. |
| 3 | The name-based version that uses MD5 hashing. Implementations SHOULD utilize UUID version 5 if possible. |
| 4 | The randomly or pseudo-randomly generated version. |
| 5 | The name-based version that uses SHA-1 hashing |
| 6 | UUID version 6 is a field-compatible version of UUIDv1, reordered for improved DB locality. It is expected that UUIDv6 will primarily be used in contexts where there are existing v1 UUIDs. Systems that do not involve legacy UUIDv1 SHOULD consider using UUIDv7 instead. |
| 7 | UUID version 7 features a time-ordered value field derived from the widely implemented and well known Unix Epoch timestamp source, the number of milliseconds seconds since midnight 1 Jan 1970 UTC, leap seconds excluded. As well as improved entropy characteristics over versions 1 or 6. Implementations SHOULD utilize UUID version 7 over UUID version 1 and 6 if possible. |
| 8 | RFC-compatible format for experimental or vendor-specific use cases. |
## 2. Test environment definition
CPU: 12th Gen Intel(R) Core(TM) i7-1255U
RAM: 16 GB
### 2.1. Database schemas
* [MySQL](docker/mysql_8_0/initdb/schema.sql)
* [Postgres](docker/postgres_15_2/initdb/schema.sql)
### 2.2. Running tests
Run next command with defined parameters:
```
docker compose run --rm php ./bin/bench [database] [benchmark] -b [batch-size] -i [iterations]
```
### 2.3 Building charts
After running benchmarks, you may build charts:
```
./bin/plot.sh
```
## 3. Results
### 3.1. Insert To Primary Key
Records inserted by batch 5000 times with size 2000. Total amount of records - 10kk.
* [MySQL 8.0 InnoDb](#mysql-80-innodb-with-promary-key)
* [Percona 8.0 InnoDb](#percona-80-innodb-with-promary-key)
* [MariaDB 10.11 InnoDb](#mariadb-1011-innodb-with-promary-key)
* [Postgres 15.2](#postgres-152-with-promary-key)
* [MongoDB 6.0](#mongodb-60-with-promary-key)
#### MySQL 8.0 InnoDb with promary key
#### Percona 8.0 InnoDb with promary key
#### MariaDB 10.11 InnoDb with promary key
#### Postgres 15.2 with promary key
#### MongoDB 6.0 with promary key
MongoDb has no auto increment feature, so external sequence used and MongoDB just insert integers.
### 3.2. Insert To Secondary Index
Records inserted by batch 5000 times with size 2000. Total amount of records - 10kk.
#### MySQL 8.0 InnoDb with secondary key
## 4. Materials
* https://www.percona.com/blog/uuids-are-popular-but-bad-for-performance-lets-discuss/
* https://www.ietf.org/rfc/rfc4122.txt
* https://uuid6.github.io/uuid6-ietf-draft/