nicecache

README

Nicecache

This is fastest golang cache*

• statically typed
• minimum read-write locks
• interface{}-free
• zero allocation for read and write

Note

I`m really need you advices and usage experience, please feel free to write me you issues, feature requests and use cases.

Usage

First of all you need to generate nicecache files. You could create a cache file in the same package as cached type:

//go:generate nicecache -type=TestValue -cacheSize=10000000
type TestValue struct {
    ...
}

Or create cache file in other package:

//go:generate nicecache -type=TestItem -cacheSize=10000000 -cachePackage=nicecache/example/repository
type TestItem struct {
    ...
}

// create and init cache
cache := NewNiceCacheTestValue()
defer cache.Close() // close when needed (app exit for example)

key := []byte("product_category_123")  // sync.Pool can be used
toStore := TestType{...}

err := cache.Set(key, &toStore, longTime)
if err != nil {
	//...
}

gotValue := TestType{} // sync.Pool can be used
err = cache.Get(key, &gotValue)
if err != nil {
	//...
}

Example

Example is here.

Benchmarks

Cache size 1.000.000 items, performance on 4 GOMAXPROCS:

Benchmark_Cache_Nice_Set-4                      100000000              132 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Set_Parallel-4             200000000               88.4 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get-4                      100000000              129 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get_Parallel-4             200000000               61.9 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet-4                100000000              121 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet_Parallel-4       100000000              166 ns/op               0 B/op          0 allocs/op
     
Benchmark_Chunked_MapStringInterface_Set-4                     2000000               894 ns/op             219 B/op          1 allocs/op
Benchmark_Chunked_MapStringInterface_Set_Parallel-4            1000000              1151 ns/op             247 B/op          1 allocs/op
Benchmark_Chunked_MapStringInterface_Get-4                     1000000              1952 ns/op             336 B/op          2 allocs/op
Benchmark_Chunked_MapStringInterface_Get_Parallel-4            1000000              2732 ns/op             336 B/op          2 allocs/op
Benchmark_Chunked_MapStringInterface_SetAndGet-4               1000000              3247 ns/op             583 B/op          3 allocs/op

Cache size 10.000.000 items, performance on 4 GOMAXPROCS:

Benchmark_Cache_Nice_Set-4                      100000000              196 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Set_Parallel-4             200000000               96.2 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get-4                      100000000              199 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get_Parallel-4             200000000               82.9 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet-4                100000000              158 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet_Parallel-4       100000000              186 ns/op               0 B/op          0 allocs/op
    
Benchmark_Chunked_MapStringInterface_Set-4                     1000000              2766 ns/op             761 B/op          5 allocs/op
Benchmark_Chunked_MapStringInterface_Set_Parallel-4            1000000              3243 ns/op             761 B/op          5 allocs/op
Benchmark_Chunked_MapStringInterface_Get-4                     1000000              1925 ns/op             336 B/op          2 allocs/op
Benchmark_Chunked_MapStringInterface_Get_Parallel-4            1000000              2345 ns/op             336 B/op          2 allocs/op
Benchmark_Chunked_MapStringInterface_SetAndGet-4                300000              5638 ns/op            1078 B/op          7 allocs/op

Cache size 1.000.000 items, performance on 8 GOMAXPROCS:

Benchmark_Cache_Nice_Set-8                      100000000              131 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Set_Parallel-8             200000000               62.4 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get-8                      100000000              131 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get_Parallel-8             300000000               42.4 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet-8                100000000              122 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet_Parallel-8       100000000              132 ns/op               0 B/op          0 allocs/op

Cache size 10.000.000 items, performance on 8 GOMAXPROCS:

Benchmark_Cache_Nice_Set-8                      100000000              191 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Set_Parallel-8             200000000               65.5 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get-8                      100000000              207 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_Get_Parallel-8             300000000               51.7 ns/op             0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet-8                100000000              155 ns/op               0 B/op          0 allocs/op
Benchmark_Cache_Nice_SetAndGet_Parallel-8       100000000              143 ns/op               0 B/op          0 allocs/op

Remarks:

*as far as i know :)