Simple, Fast, In-Memory Key-Value store.
- In-memory store for very quick access
- Basic Operations: PUT, GET,
DELETE - Exposed via REST API
Potential Future Features:
- REPL for interacting with the store
- Metrics, accessible via the REPL
git clone https://github.com/podikoglou/smolkey
cd smolkey
go run cmd/main.go
The Smolkey API is now running on port 8080.
$ curl -X PUT -d value=bar http://localhost:8080/foo
Created
$ curl http://localhost:8080/foo
bar
import requests
requests.put("http://localhost:8080/foo", data={"value": "bar"})
response = requests.get("http://localhost:8080/foo")
print(response.text) # => barAll benchmarks were performed on an M1 Mac using wrk with 6 threads and 60 connections.
Writing Random Keys with Random Values:
Thread Stats Avg Stdev Max +/- Stdev
Latency 11.79ms 32.21ms 251.35ms 89.68%
Req/Sec 36.05k 28.12k 159.33k 76.73%
2030546 requests in 10.10s, 247.87MB read
Requests/sec: 201094.97
Transfer/sec: 24.55MB
Reading Random Keys:
Thread Stats Avg Stdev Max +/- Stdev
Latency 12.28ms 32.68ms 269.33ms 89.47%
Req/Sec 37.82k 30.49k 179.97k 75.25%
2103317 requests in 10.05s, 243.71MB read
Non-2xx or 3xx responses: 59431
Requests/sec: 209377.07
Transfer/sec: 24.26MB
Reading Specific Key:
Thread Stats Avg Stdev Max +/- Stdev
Latency 11.59ms 32.35ms 256.85ms 90.28%
Req/Sec 46.51k 40.60k 209.97k 74.56%
2626620 requests in 10.09s, 298.09MB read
Requests/sec: 260349.99
Transfer/sec: 29.55MB
It's a toy project.
Smolkey follows a philosophy similar to that of Redis. We trust you to run Smolkey under a firewall, so we really don't need to implement authentication in the application level. (if you really want authentication, you can theoretically proxy it through something like Caddy.)
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent