sniper

GoDoc

A simple and efficient thread-safe key/value store for Go.

NOTES: Original sniper repositroy not support Bucket (sortedset) reindex. If database closed, bucket indexes pufff. This repository supports:

• Bucket reindexing after database opened or chucks init.
• Add new Index methods: Remove, PutIndex, RemoveIndex, HasIndex -> Look at the tests file how to use it

TODO: Bucket Items count method is missing!!!

Getting Started

Features

• Store hundreds of millions of entries
• Fast. High concurrent. Thread-safe. Scales on multi-core CPUs
• Extremly low memory usage
• Zero GC overhead
• Simple, pure Go implementation

Installing

To start using sniper, install Go and run go get:

Original Package

$ go get -u github.com/recoilme/sniper

Modify Package

$ go get -u github.com/uretgec/sniper

This will retrieve the library.

Usage

The Sniper includes this methods: Set, Get, Incr, Decr, Delete, Count, Open, Close, FileSize, Backup.

s, _ := sniper.Open(sniper.Dir("1"))
s.Set([]byte("hello"), []byte("go"))
res, _ = s.Get([]byte("hello"))
fmt.Println(res)
s.Close()
// Output:
// go

with Bucket includes this methods: (this methods can use only this repository) CreateBucketIndexIfNotExists,Put,Remove,Keys,PutIndex,RemoveIndex,HasIndex.

s, _ := sniper.Open(sniper.Dir("2"),CreateBucketIndexIfNotExists([]string{"users"}))

users, _ := s.Bucket("users")
s.Set([]byte("users01")), []byte("rob"), 0) // Set users01 -> rob
s.PutIndex(users, []byte("01")) // users01 put into users bucket index

keys := users.Keys(0, 0)
fmt.Printf("%v#\n", keys) // []string{"01"}

has := s.HasIndex(users, []byte("01"))
fmt.Println(has) // true

deleted := s.RemoveIndex(users, []byte("02"))
fmt.Println(deleted) // false

res, _ = s.Get([]byte("users01"))
fmt.Println(res)

s.Close()
// Output:
// go

Performance

MacBook Pro 2019 (Quad-Core Intel Core i7 2,8 GHz, 16 ГБ, APPLE SSD AP0512M)

go version go1.14 darwin/amd64

     number of cpus: 8
     number of keys: 10000000
            keysize: 10
        random seed: 1570109110136449000

-- sniper --
set: 10,000,000 ops over 8 threads in 63159ms, 158,331/sec, 6315 ns/op, 644.3 MB, 67 bytes/op
get: 10,000,000 ops over 8 threads in 4455ms, 2,244,629/sec, 445 ns/op, 305.5 MB, 32 bytes/op
del: 10,000,000 ops over 8 threads in 37568ms, 266,182/sec, 3756 ns/op, 122.8 MB, 12 bytes/op

With fsync

set: 10,000,000 ops over 8 threads in 85088ms, 117,524/sec, 8508 ns/op, 644.4 MB, 67 bytes/op
get: 10,000,000 ops over 8 threads in 5623ms, 1,778,268/sec, 562 ns/op, 305.5 MB, 32 bytes/op

How it is done

• Sniper database is sharded on 250+ chunks. Each chunk has its own lock (RW), so it supports high concurrent access on multi-core CPUs.
• Each chunk store hash(key) -> (value addr, value size), map.
• Hash is very short, and has collisions. Sniper has collisions resolver.
• Efficient space reuse alghorithm. Every packet has power of 2 size, for inplace rewrite on value update and map of deleted entrys, for reusing space.

Limitations

• 512 Kb - maximum entry size len(key) + len(value)
• ~1 Tb - maximum database size

Mac OS tip

How to Change Open Files Limit on OS X and macOS

Contact

Vadim Kulibaba @recoilme

License

sniper source code is available under the MIT License.