Crocdb ^new^ -
| Operation | Throughput (ops/sec) | Latency (p99) | |-----------|----------------------|---------------| | SET (No fsync) | 1,200,000 | 0.15 ms | | SET (Periodic fsync) | 350,000 | 0.5 ms | | GET | 1,800,000 | 0.08 ms | | DEL | 1,100,000 | 0.2 ms |
| Command | Description | Example | |---------|-------------|---------| | SET key value [TTL seconds] | Stores a value. | SET user:100 "john" TTL 60 | | GET key | Retrieves a value. | GET user:100 | | DEL key [key...] | Deletes one or more keys. | DEL user:100 user:101 | | EXISTS key | Returns 1 if key exists. | EXISTS user:100 | | KEYS pattern | Lists keys matching a glob pattern (scan-only, O(N)). | KEYS user:* | | FLUSH | Clears all data. | FLUSH | | SAVE | Triggers an immediate snapshot. | SAVE | | PING | Health check. | PING | crocdb
func main() db, err := croc.Open(croc.Options Path: "/data/croc", FSyncMode: croc.Periodic, ) if err != nil log.Fatal(err) | Operation | Throughput (ops/sec) | Latency (p99)
package main import ( "fmt" "log" "github.com/croc-db/croc" ) | DEL user:100 user:101 | | EXISTS key
If your workload fits entirely in RAM and you need restart safety without operating a full database cluster, CrocDB is an excellent choice. For the latest documentation, source code, and benchmarks, visit the official repository: https://github.com/croc-db/croc
defer db.Close()
1. Overview & Definition CrocDB is an open-source, lightweight, and high-performance key-value database written in Go. Its name is a playful nod to "Redis" (like the reptile) while emphasizing its distinct architectural choices. CrocDB is designed for developers who need a fast, embedded database that combines the low-latency benefits of an in-memory store with the durability of on-disk persistence.
