https://github.com/vladpodilnyk/smoothie
A rate limiter that smooth your day!
https://github.com/vladpodilnyk/smoothie
distributed-systems rate-limiter redis
Last synced: 4 months ago
JSON representation
A rate limiter that smooth your day!
- Host: GitHub
- URL: https://github.com/vladpodilnyk/smoothie
- Owner: VladPodilnyk
- License: mit
- Created: 2020-02-20T15:52:25.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2024-02-14T23:21:33.000Z (over 1 year ago)
- Last Synced: 2024-06-21T06:44:03.797Z (over 1 year ago)
- Topics: distributed-systems, rate-limiter, redis
- Language: Go
- Homepage:
- Size: 16.6 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Smoothie
A dead simple distributed rate limiter that smoothes your day.
### Yet another rate limiter?
Smoothie is meant to be a distributed rate limiter, unlike all other popular within go community
solutions. The project is quite small so, it doesn't make
any sense to publish it as a package. But it can be a good showcase of how to
make a distributed rate limiter using redis and Go.
If you don't need that or maybe you don't like Redis,
then standard `time` package or `rate` package might serve you better.
### Implementation details
At the moment smoothie implements a fixed-size window strategy for limiting requests.
The api resembles the familiar `rate` package API with `Allow` and `Exec` methods.
See example below
```go
config := redis.Options{
Addr: "localhost:6379",
Password: "",
DB: 0,
}
limiter := New(&testRedisOptions, Rate{NumberOfRequests: 1, Duration: 5 * time.Second})
isAllowed := limiter.Allow(ctx.Background(), "user-IP-address")
testEffect := func () error {
fmt.Println("hello")
return nil
}
err := limiter.Exec(ctx.Background(), "user-IP-adress", testEffect)
```
### Future plans
It would be nice to add the following things to the smoothie:
- other rate limiting strategies (Sliding window, token bucket)
- extend tests with test cases with concurrent access to the same limiting `key`