fastcache

A fast, generic, thread-safe cache for Go.

Note

This is a fork of VictoriaMetrics/fastcache with a redesigned API using Go generics.

Features

Generic: Type-safe API.
Zero allocations: No allocations on Get ops.
Thread-safe: Concurrent goroutines may read and write into a single cache instance.
FIFO eviction: Oldest entries are evicted first when the cache is full.
Iterators: Go 1.23+ range-over-func support with All(), Keys(), Values().
Atomic operations: GetOrSet, GetAndDelete, SetIfAbsent for lock-free patterns.
Persistence: Cache can be saved to file and loaded from file.
Simple API: See Go reference.

Install

go get go.dw1.io/fastcache

Usage

package main

import (
    "fmt"

    "go.dw1.io/fastcache"
)

func main() {
    // create a cache with max 10000 entries
    c, err := fastcache.New[string, int](10000)
    if err != nil {
        fmt.Println("create cache:", err)
        return
    }

    // set values
    if err := c.Set("foo", 123); err != nil {
        fmt.Println("set foo:", err)
        return
    }
    if err := c.Set("bar", 456); err != nil {
        fmt.Println("set bar:", err)
        return
    }

    // get values
    if v, ok := c.Get("foo"); ok {
        fmt.Println("foo =", v) // foo = 123
    }

    // check existence
    if c.Has("bar") {
        fmt.Println("bar exists")
    }

    // delete
    c.Delete("foo")

    // iterate
    for k, v := range c.All() {
        fmt.Printf("%s: %d\n", k, v)
    }

    // iterate keys only
    for k := range c.Keys() {
        fmt.Println(k)
    }

    // atomic get-or-set
    actual, loaded, err := c.GetOrSet("baz", 789)
    if err != nil {
        fmt.Println("get or set baz:", err)
        return
    }
    fmt.Printf("value=%d, existed=%v\n", actual, loaded)

    // atomic set-if-absent
    stored, err := c.SetIfAbsent("qux", 101112)
    if err != nil {
        fmt.Println("set if absent qux:", err)
        return
    }
    if stored {
        fmt.Println("qux was set")
    }

    // atomic get-and-delete
    if v, ok := c.GetAndDelete("bar"); ok {
        fmt.Println("deleted bar:", v)
    }

    // get stats
    var stats fastcache.Stats
    c.UpdateStats(&stats)
    fmt.Printf("Hits: %d, Misses: %d\n", stats.Hits, stats.Misses)

    // save to file
    if err := c.SaveToFile("/tmp/cache.bin"); err != nil {
	    fmt.Println("save cache:", err)
	    return
    }

    // load from file
    c2, err := fastcache.LoadFromFile[string, int]("/tmp/cache.bin")
    if err != nil {
	    fmt.Println("load cache:", err)
	    return
    }
    defer c2.Reset()
}

Architecture

The cache uses a sharded design for concurrent scalability:

512 shards: Each with its own lock, reducing contention on multi-core CPUs.
Generic map storage: map[K]V per shard for O(1) lookups.
Ring buffer for FIFO: Circular buffer tracks insertion order for eviction.

Differences from Original

Aspect	Original fastcache	This fork
API	`[]byte` keys/values	Generic `[K, V]`
Capacity	Bytes-based	Entry-count based
Storage	Ring buffer of bytes	`map[K]V`
Eviction	FIFO (by byte position)	FIFO (by insertion order)
Allocations	1 alloc/Get	Zero

Benchmarks

Compared against the original VictoriaMetrics/fastcache and maypok86/otter:

benchstat

goos: linux
goarch: amd64
pkg: benchmarks
cpu: AMD EPYC 7763 64-Core Processor                
                  │ fastcache_fork │               fastcache               │                 otter                  │
                  │     sec/op     │    sec/op     vs base                 │    sec/op     vs base                  │
Set/1-4                20.37n ±  0%    44.07n ± 4%  +116.40% (p=0.000 n=10)   233.15n ± 1%  +1044.86% (p=0.000 n=10)
Get/1-4                18.26n ±  0%    46.78n ± 0%  +156.16% (p=0.000 n=10)    33.83n ± 0%    +85.30% (p=0.000 n=10)
SetGet/1-4             37.48n ±  0%    88.45n ± 0%  +135.98% (p=0.000 n=10)   278.35n ± 1%   +642.66% (p=0.000 n=10)
GetParallel/1-4        43.91n ±  1%    48.23n ± 0%    +9.84% (p=0.000 n=10)    15.38n ± 1%    -64.98% (p=0.000 n=10)
SetParallel/1-4        46.76n ±  2%    66.33n ± 1%   +41.88% (p=0.000 n=10)   251.80n ± 1%   +438.55% (p=0.000 n=10)
Set/16-4               22.29n ±  0%    41.59n ± 2%   +86.65% (p=0.000 n=10)   237.90n ± 2%   +967.53% (p=0.000 n=10)
Get/16-4               19.91n ±  0%    53.87n ± 2%  +170.57% (p=0.000 n=10)    35.33n ± 1%    +77.42% (p=0.000 n=10)
SetGet/16-4            41.35n ±  0%    93.64n ± 1%  +126.44% (p=0.000 n=10)   279.05n ± 1%   +574.85% (p=0.000 n=10)
GetParallel/16-4       46.31n ±  0%    46.90n ± 0%    +1.26% (p=0.000 n=10)    15.84n ± 0%    -65.80% (p=0.000 n=10)
SetParallel/16-4       51.38n ±  3%    70.65n ± 1%   +37.50% (p=0.000 n=10)   255.00n ± 0%   +396.30% (p=0.000 n=10)
Set/32-4               26.00n ±  0%    49.40n ± 0%   +89.98% (p=0.000 n=10)   238.75n ± 1%   +818.09% (p=0.000 n=10)
Get/32-4               23.80n ±  0%    62.97n ± 0%  +164.61% (p=0.000 n=10)    36.14n ± 1%    +51.88% (p=0.000 n=10)
SetGet/32-4            49.09n ±  0%   108.10n ± 0%  +120.21% (p=0.000 n=10)   281.80n ± 1%   +474.05% (p=0.000 n=10)
GetParallel/32-4       34.55n ±  2%    41.28n ± 2%   +19.49% (p=0.000 n=10)    16.50n ± 0%    -52.24% (p=0.000 n=10)
SetParallel/32-4       39.73n ± 10%    61.83n ± 1%   +55.59% (p=0.000 n=10)   232.40n ± 3%   +484.87% (p=0.000 n=10)
Set/128-4              31.45n ±  0%    63.14n ± 2%  +100.75% (p=0.000 n=10)   242.05n ± 1%   +669.63% (p=0.000 n=10)
Get/128-4              29.32n ±  0%    96.62n ± 1%  +229.46% (p=0.000 n=10)    38.24n ± 0%    +30.42% (p=0.000 n=10)
SetGet/128-4           59.61n ±  0%   151.55n ± 0%  +154.21% (p=0.000 n=10)   295.50n ± 1%   +395.68% (p=0.000 n=10)
GetParallel/128-4      16.50n ±  2%    65.49n ± 1%  +296.79% (p=0.000 n=10)    18.68n ± 0%    +13.18% (p=0.000 n=10)
SetParallel/128-4      24.03n ± 11%    29.95n ± 8%   +24.64% (p=0.000 n=10)   240.90n ± 3%   +902.50% (p=0.000 n=10)
Set/256-4              40.51n ±  0%    79.31n ± 1%   +95.82% (p=0.000 n=10)   247.30n ± 1%   +510.54% (p=0.000 n=10)
Get/256-4              38.63n ±  1%   131.10n ± 1%  +239.37% (p=0.000 n=10)    51.40n ± 0%    +33.06% (p=0.000 n=10)
SetGet/256-4           77.30n ±  0%   215.10n ± 0%  +178.27% (p=0.000 n=10)   300.95n ± 1%   +289.33% (p=0.000 n=10)
GetParallel/256-4      20.40n ±  8%    78.72n ± 2%  +285.88% (p=0.000 n=10)    22.98n ± 0%    +12.67% (p=0.000 n=10)
SetParallel/256-4      24.67n ±  9%    39.66n ± 5%   +60.77% (p=0.000 n=10)   248.85n ± 1%   +908.92% (p=0.000 n=10)
Set/512-4              70.83n ±  0%   110.30n ± 0%   +55.74% (p=0.000 n=10)   258.20n ± 0%   +264.56% (p=0.000 n=10)
Get/512-4              70.82n ±  1%   168.10n ± 1%  +137.36% (p=0.000 n=10)    58.24n ± 0%    -17.76% (p=0.000 n=10)
SetGet/512-4           139.0n ±  0%    271.1n ± 0%   +95.04% (p=0.000 n=10)    319.6n ± 0%   +129.89% (p=0.000 n=10)
GetParallel/512-4      30.91n ±  5%   100.40n ± 2%  +224.87% (p=0.000 n=10)    25.65n ± 0%    -17.00% (p=0.000 n=10)
SetParallel/512-4      34.40n ±  6%    44.66n ± 3%   +29.81% (p=0.000 n=10)   250.15n ± 0%   +627.18% (p=0.000 n=10)
Set/1024-4             97.08n ±  0%   175.70n ± 0%   +80.98% (p=0.000 n=10)   278.65n ± 1%   +187.02% (p=0.000 n=10)
Get/1024-4             97.04n ±  0%   317.75n ± 3%  +227.44% (p=0.000 n=10)    73.27n ± 1%    -24.50% (p=0.000 n=10)
SetGet/1024-4          191.2n ±  0%    437.9n ± 0%  +129.09% (p=0.000 n=10)    340.6n ± 0%    +78.18% (p=0.000 n=10)
GetParallel/1024-4     42.58n ±  0%   161.60n ± 1%  +279.52% (p=0.000 n=10)    33.59n ± 0%    -21.11% (p=0.000 n=10)
SetParallel/1024-4     44.45n ±  2%    74.70n ± 1%   +68.07% (p=0.000 n=10)   245.60n ± 1%   +452.53% (p=0.000 n=10)
Set/2048-4             161.6n ±  0%    338.5n ± 0%  +109.47% (p=0.000 n=10)    330.7n ± 1%   +104.64% (p=0.000 n=10)
Get/2048-4             160.3n ±  0%    669.9n ± 2%  +317.90% (p=0.000 n=10)    126.3n ± 2%    -21.21% (p=0.000 n=10)
SetGet/2048-4          307.8n ±  0%    816.8n ± 1%  +165.41% (p=0.000 n=10)    415.6n ± 1%    +35.06% (p=0.000 n=10)
GetParallel/2048-4     72.24n ±  2%   305.70n ± 1%  +323.17% (p=0.000 n=10)    52.02n ± 0%    -27.98% (p=0.000 n=10)
SetParallel/2048-4     73.79n ±  1%   134.05n ± 1%   +81.66% (p=0.000 n=10)   227.75n ± 1%   +208.65% (p=0.000 n=10)
Set/4096-4             310.0n ±  1%    649.1n ± 0%  +109.39% (p=0.000 n=10)    519.1n ± 1%    +67.44% (p=0.000 n=10)
Get/4096-4             308.3n ±  1%   1068.5n ± 1%  +246.58% (p=0.000 n=10)    284.9n ± 0%     -7.59% (p=0.000 n=10)
SetGet/4096-4          587.5n ±  0%   1629.5n ± 0%  +177.34% (p=0.000 n=10)    655.1n ± 1%    +11.51% (p=0.000 n=10)
GetParallel/4096-4     131.1n ±  1%    526.8n ± 1%  +301.79% (p=0.000 n=10)    102.0n ± 3%    -22.20% (p=0.000 n=10)
SetParallel/4096-4     133.7n ±  1%    254.6n ± 1%   +90.43% (p=0.000 n=10)    263.2n ± 3%    +96.86% (p=0.000 n=10)
Set/8192-4             578.6n ±  1%   1204.0n ± 0%  +108.09% (p=0.000 n=10)    759.1n ± 2%    +31.20% (p=0.000 n=10)
Get/8192-4             573.5n ±  1%   1907.5n ± 4%  +232.64% (p=0.000 n=10)    515.1n ± 1%    -10.17% (p=0.000 n=10)
SetGet/8192-4         1072.0n ±  0%   3257.0n ± 1%  +203.82% (p=0.000 n=10)    965.4n ± 2%     -9.94% (p=0.000 n=10)
GetParallel/8192-4     244.8n ±  0%    851.1n ± 1%  +247.74% (p=0.000 n=10)    188.2n ± 5%    -23.08% (p=0.000 n=10)
SetParallel/8192-4     247.3n ±  0%    513.6n ± 3%  +107.72% (p=0.000 n=10)    375.2n ± 1%    +51.77% (p=0.000 n=10)
geomean                69.81n          158.8n       +127.46%                   149.3n        +113.91%

                  │ fastcache_fork │            fastcache             │              otter               │
                  │      B/op      │     B/op      vs base            │     B/op      vs base            │
Set/1-4                 0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/1-4                0.000 ± 0%       8.000 ± 0%  ? (p=0.000 n=10)       0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/1-4             0.000 ± 0%       8.000 ± 0%  ? (p=0.000 n=10)      64.000 ± 0%  ? (p=0.000 n=10)
GetParallel/1-4        0.000 ± 0%       8.000 ± 0%  ? (p=0.000 n=10)       0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/1-4         0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     65.00 ± 0%  ? (p=0.000 n=10)
Set/16-4                0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/16-4                0.00 ± 0%       16.00 ± 0%  ? (p=0.000 n=10)        0.00 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/16-4             0.00 ± 0%       16.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/16-4        0.00 ± 0%       16.00 ± 0%  ? (p=0.000 n=10)        0.00 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/16-4        0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     65.00 ± 0%  ? (p=0.000 n=10)
Set/32-4                0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/32-4                0.00 ± 0%       32.00 ± 0%  ? (p=0.000 n=10)        0.00 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/32-4             0.00 ± 0%       32.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/32-4        0.00 ± 0%       32.00 ± 0%  ? (p=0.000 n=10)        0.00 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/32-4        0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/128-4               0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/128-4                0.0 ± 0%       128.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/128-4            0.00 ± 0%      128.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/128-4        0.0 ± 0%       128.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/128-4       0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/256-4               0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/256-4                0.0 ± 0%       234.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/256-4            0.00 ± 0%      256.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/256-4        0.0 ± 0%       234.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/256-4       0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/512-4               0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/512-4                0.0 ± 0%       490.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/512-4            0.00 ± 0%      512.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/512-4        0.0 ± 0%       490.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/512-4       0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/1024-4              0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/1024-4               0.0 ± 0%       938.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/1024-4           0.00 ± 0%     1024.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/1024-4       0.0 ± 0%       938.0 ± 0%  ? (p=0.000 n=10)         0.0 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/1024-4      0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/2048-4              0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/2048-4           0.000Ki ± 0%     1.749Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetGet/2048-4           0.00 ± 0%     2048.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/2048-4   0.000Ki ± 0%     1.749Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/2048-4      0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/4096-4              0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/4096-4           0.000Ki ± 0%     3.243Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetGet/4096-4           0.00 ± 0%     4096.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/4096-4   0.000Ki ± 0%     3.239Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/4096-4      0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Set/8192-4              0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
Get/8192-4           0.000Ki ± 0%     6.478Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetGet/8192-4           0.00 ± 0%     8192.00 ± 0%  ? (p=0.000 n=10)       64.00 ± 0%  ? (p=0.000 n=10)
GetParallel/8192-4   0.000Ki ± 0%     6.474Ki ± 0%  ? (p=0.000 n=10)     0.000Ki ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/8192-4      0.00 ± 0%        0.00 ± 0%  ~ (p=1.000 n=10) ¹     64.00 ± 0%  ? (p=0.000 n=10)
geomean                           ²                 ?                ²                 ?                ²
¹ all samples are equal
² summaries must be >0 to compute geomean

                  │ fastcache_fork │           fastcache            │             otter              │
                  │   allocs/op    │ allocs/op   vs base            │ allocs/op   vs base            │
Set/1-4                0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/1-4                0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/1-4             0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/1-4        0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/1-4        0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/16-4               0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/16-4               0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/16-4            0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/16-4       0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/16-4       0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/32-4               0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/32-4               0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/32-4            0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/32-4       0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/32-4       0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/128-4              0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/128-4              0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/128-4           0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/128-4      0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/128-4      0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/256-4              0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/256-4              0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/256-4           0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/256-4      0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/256-4      0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/512-4              0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/512-4              0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/512-4           0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/512-4      0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/512-4      0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/1024-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/1024-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/1024-4          0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/1024-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/1024-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/2048-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/2048-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/2048-4          0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/2048-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/2048-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/4096-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/4096-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/4096-4          0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/4096-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/4096-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Set/8192-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
Get/8192-4             0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetGet/8192-4          0.000 ± 0%     1.000 ± 0%  ? (p=0.000 n=10)     1.000 ± 0%  ? (p=0.000 n=10)
GetParallel/8192-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   0.000 ± 0%  ~ (p=1.000 n=10) ¹
SetParallel/8192-4     0.000 ± 0%     0.000 ± 0%  ~ (p=1.000 n=10) ¹   1.000 ± 0%  ? (p=0.000 n=10)
geomean                           ²               ?                ²               ?                ²
¹ all samples are equal
² summaries must be >0 to compute geomean

Highlights:

~2x+ faster geomean sec/op vs original fastcache and otter.
Zero B/op and zero allocations in measured ops; original fastcache still shows non-zero B/op across many read and mixed workloads, while otter incurs B/op and allocations on writes.
Otter still wins several read-heavy micro-benchmarks, especially GetParallel and some larger-key Get cases; this fork leads decisively on writes and most mixed workloads.

Run benchmarks yourself:

make bench -C benchmarks

Limitations

No cache expiration: entries are evicted only when the cache is full (FIFO order).
No size-based limits: capacity is by entry count, not bytes.

Status

Caution

fastcache is pre-v1 and does NOT provide a stable API; use at your own risk.

Occasional breaking changes may be introduced without notice until a post-v1 release.

License

MIT. Same as original. See LICENSE.

Name		Name	Last commit message	Last commit date
Latest commit History 96 Commits
.github/workflows		.github/workflows
benchmarks		benchmarks
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
cache.go		cache.go
cache_example_test.go		cache_example_test.go
cache_test.go		cache_test.go
cache_timing_test.go		cache_timing_test.go
doc.go		doc.go
errors.go		errors.go
file.go		file.go
file_example_test.go		file_example_test.go
file_test.go		file_test.go
go.mod		go.mod
go.sum		go.sum
shard.go		shard.go
stats.go		stats.go

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

fastcache

Features

Install

Usage

Architecture

Differences from Original

Benchmarks

Limitations

Status

License

About

Uh oh!

Contributors

Languages

Folders and files

Latest commit

History

Repository files navigation

fastcache

Features

Install

Usage

Architecture

Differences from Original

Benchmarks

Limitations

Status

License

About

Topics

Resources

License

Uh oh!

Stars

Watchers

Forks

Contributors 0

Languages

Contributors