High-performance database. Basically network and disk synchronized hashmaps.
BenchmarkCollectionGet-12 317030264 3.75 ns/op 0 B/op 0 allocs/op
BenchmarkCollectionSet-12 11678318 102 ns/op 32 B/op 2 allocs/op
BenchmarkCollectionDelete-12 123748969 9.50 ns/op 5 B/op 0 allocs/op
BenchmarkCollectionAll-12 1403905 859 ns/op 2144 B/op 2 allocs/op
- Low latency commands
- Every command is "local first, sync later"
- Data is stored in memory
- Data is synchronized between all nodes in a cluster
- Data is saved to disk persistently using JSON
- Timestamp based conflict resolution
- Uses the extremely fast
sync.Map
- Namespace: Contains multiple collections (e.g. "google")
- Collection: Contains homogeneous data for a data type (e.g. "User")
- Key: The string that lets you look up a single object in a collection
All of the above require a unique name. Given namespace, collection and key, you can access the data stored for it.
Please take a look at the style guidelines if you'd like to make a pull request.
Cedric Fung | Scott Rayapoullé | Eduard Urbach |
Want to see your own name here?