1 unstable release
0.1.0 | Nov 25, 2023 |
---|
#1429 in Asynchronous
29KB
465 lines
GrausDb
GrausDb is a high-performance, thread-safe key-value embedded data store written in Rust. It is designed for simplicity, efficiency, and reliability.
Features
- Lock-Free Concurrency: GrausDb leverages lock-free data structures for high-performance concurrent access.
- Persistence: Data is persisted to disk for durability.
- Log-Based Storage: Key-value pairs are stored in log files.
- Benchmarks: Benchmarks included to assess the performance.
Installation
To use GrausDb in your Rust project, simply add it as a dependency in your Cargo.toml
file:
[dependencies]
graus_db = "0.1.0"
Quick start
Here's a quick example of how to use GrausDb in your Rust application:
let store = GrausDb::open("path")?;
store.set("key".to_owned(), "value".to_owned())?;
let val = store.get("key".to_owned())?;
It can also be called from multiple threads:
let store = GrausDb::open("path")?;
// Calls set method from 8 different threads
for i in 0..8 {
let store = store.clone();
thread::spawn(move || {
store.set(format!("key{}", i), format!("value{}", i)).unwrap();
});
}
API
GrausDb provides a simple and intuitive API for interacting with the key-value store. Below are some of the key functions and methods exposed by GrausDb, along with usage examples.
GrausDb::open
open
is used to open a GrausDb instance, creating a new database if it doesn't exist at the specified path.
Example:
use graus_db::{GrausDb, Result};
fn main() -> Result<()> {
let store = GrausDb::open("my_database")?;
// Your database is now ready to use.
Ok(())
}
set
The set
method is used to store a key-value pair in the database.
Example:
use graus_db::{GrausDb, Result};
fn main() -> Result<()> {
let store = GrausDb::open("my_database")?;
store.set("key".to_owned(), "value".to_owned())?;
// Key "key" now has the value "value" in the database.
Ok(())
}
get
The get
method retrieves the value associated with a given key.
Example:
use graus_db::{GrausDb, Result};
fn main() -> Result<()> {
let store = GrausDb::open("my_database")?;
store.set("key".to_owned(), "value".to_owned())?;
if let Some(value) = store.get("key".to_owned())? {
println!("Value: {}", value); // Outputs: "Value: value"
} else {
println!("Key not found");
}
Ok(())
}
remove
The remove
method deletes a key and its associated value from the database.
Example:
use graus_db::{GrausDb, Result};
fn main() -> Result<()> {
let store = GrausDb::open("my_database")?;
store.set("key".to_owned(), "value".to_owned())?;
store.remove("key".to_owned())?;
// Key "key" and its value are now removed from the database.
Ok(())
}
update_if
The update_if
method updates the value of an existing key atomically, allowing you to provide a custom update function.
An optional predicate can be passed, the value will only be updated if the predicate is satisfied.
Example:
use graus_db::{GrausDb, Result};
fn main() -> Result<()> {
let key = ¨key1¨;
store.set(key.to_owned(), "25".to_owned()).unwrap();
let update_fn = |value: String| {
let num = value.parse::<i32>().unwrap();
(num - 1).to_string()
};
let predicate = |value: String| {
let num = value.parse::<i32>().unwrap();
num > 0
};
let result = store.update_if(
key.to_owned(),
update_fn,
Some(key.to_owned()),
Some(predicate),
);
// Key "key1" now has the value "24" in the database.
// The function was applied because the predicate was met (25 > 0)
}
For more details on how to use GrausDb, please refer to the tests.
Architecture and Implementation
GrausDb's architecture is built around the principles of log-based storage and lock-free concurrency:
-
Log-Based Storage: GrausDb stores key-value pairs in log files. Log files are named after monotonically increasing generation numbers with a log extension. This design ensures that data is durably persisted to disk.
-
Lock-Free Concurrency: GrausDb uses lock-free data structures to provide high-performance concurrent access to the data. This enables multiple threads to interact with the database efficiently.
-
In-Memory Index: GrausDb maintains an in-memory index that maps keys to their positions in the log. This index allows for fast lookups and efficient data retrieval.
-
Compaction: To maintain efficient storage and reduce disk space usage, GrausDb performs compaction when a threshold is reached. Compaction involves rewriting log files, removing stale data, and reclaiming disk space.
Benchmarks
GrausDb includes built-in benchmarking tools to evaluate its efficiency and to help you make data-driven decisions.
Future Development
Next features:
- Multithread benchmark
- Range get
- Sync API (journal)
- Internal threadpool + futures
License
GrausDb is licensed under the MIT License.
Happy coding with GrausDb!
Contact
GrausDb is created and maintained by Ricardo Pallas.
Website: https://rpallas92.github.io/
Dependencies
~1.1–2.1MB
~42K SLoC