Ember 是一個先進的多層快取系統，設計用於提升應用程式的效能和響應速度。它結合了本地快取和分散式快取的優勢，提供高效能、低延遲的數據存取能力。Ember 支持多種先進功能，如布隆過濾器（Bloom Filter）、自適應 TTL（Time-to-Live）、預取（Prefetching）功能等，使其成為處理高並發數據訪問和大規模數據存取的理想選擇。
Ember 提供了一個靈活且可擴展的解決方案，適用於各種使用場景，包括但不限於：
- Web 應用程式的數據快取
- API 響應優化
- 分散式系統的數據同步
- 多層快取系統：支持本地快取（Local Cache）和遠端快取（Remote Cache），提供高效能的數據存取能力。
- 布隆過濾器支持：使用布隆過濾器來高效檢測數據存在性，減少對遠端快取的訪問次數。
- 自適應 TTL：根據數據的訪問頻率動態調整 TTL，有效提高數據的命中率和快取使用率。
- 預取功能：支持預取功能，根據使用模式主動加載數據，進一步減少延遲。
- 高效的熔斷器和重試機制：內建熔斷器和重試機制，提升系統穩定性和容錯能力。
- 可配置性強：支持多種配置選項，適應不同的應用需求和場景。
- 輕量化依賴：依賴最小化，易於集成和部署。
go get -u goflare.io/ember以下是一個簡單的使用範例：
package main
import (
"context"
"fmt"
"log"
"time"
"github.com/redis/go-redis/v9"
"go.uber.org/zap"
"goflare.io/ember"
)
func main() {
// 創建上下文
ctx := context.Background()
// 初始化 Logger
logger, err := zap.NewDevelopment()
if err != nil {
log.Fatalf("Failed to initialize logger: %v", err)
}
defer logger.Sync()
// 配置 Redis 選項
redisOptions := &redis.Options{
Addr: "localhost:6379", // 根據實際情況設置
Password: "", // 無密碼
DB: 0, // 默認數據庫
}
// 初始化 Ember 庫
emberCache, err := ember.New(
ctx,
redisOptions,
ember.WithLogger(logger),
ember.WithMaxLocalSize(2*1024*1024*1024), // 2GB
ember.WithShardCount(16),
ember.WithDefaultExpiration(10*time.Minute),
ember.WithSerialization("json"),
)
if err != nil {
log.Fatalf("Failed to initialize Ember: %v", err)
}
defer func() {
if err := emberCache.Close(); err != nil {
logger.Error("Failed to close Ember", zap.Error(err))
}
}()
// 設置快取項目
key := "example_key"
value := "Hello, Ember!"
ttl := 5 * time.Minute
if err := emberCache.Set(ctx, key, value, ttl); err != nil {
log.Fatalf("Set failed: %v", err)
}
fmt.Printf("Set key '%s' with value '%s' and TTL %v\n", key, value, ttl)
// 獲取快取項目
var retrieved string
found, err := emberCache.Get(ctx, key, &retrieved)
if err != nil {
log.Fatalf("Get failed: %v", err)
}
if found {
fmt.Printf("Retrieved key '%s' with value '%s'\n", key, retrieved)
} else {
fmt.Printf("Key '%s' not found\n", key)
}
// 刪除快取項目
if err := emberCache.Delete(ctx, key); err != nil {
log.Fatalf("Delete failed: %v", err)
}
fmt.Printf("Deleted key '%s'\n", key)
// 清空所有快取項目
if err := emberCache.Clear(ctx); err != nil {
log.Fatalf("Clear failed: %v", err)
}
fmt.Println("Cleared all cache entries")
}Ember 提供了豐富的配置選項，以下是主要配置參數的詳細說明：
- EnableLocalCache: 是否啟用本地快取，默認為
true。 - MaxLocalSize: 本地快取的最大大小，以字節為單位。
- ShardCount: 本地快取的分片數量，建議設置為 CPU 核心數的 2 到 4 倍。
- DefaultExpiration: 默認的數據過期時間。
- CleanupInterval: 清理過期項目的時間間隔。
- CacheBehaviorConfig: 包含預取（Prefetching）、自適應 TTL（Adaptive TTL）、布隆過濾器（Bloom Filter）等行為配置。
- ResilienceConfig: 包含熔斷器和重試機制的配置。
- Serialization: 指定數據的序列化和反序列化方式。
歡迎任何形式的貢獻！請參閱 CONTRIBUTING.md 了解更多信息。
Ember 根據 MIT 許可證分發。詳細信息請參閱 LICENSE。
Ember is an advanced multi-level caching system designed to enhance application performance and responsiveness. It combines the benefits of local and distributed caching, providing efficient and low-latency data access capabilities. Ember supports a variety of advanced features such as Bloom Filter support, Adaptive TTL, and Prefetching, making it ideal for handling high-concurrency data access and large-scale data storage scenarios.
Ember provides a flexible and scalable solution suitable for various use cases, including but not limited to:
- Data caching for web applications
- API response optimization
- Data synchronization in distributed systems
- Multi-Level Caching System: Supports both Local Cache and Remote Cache, providing high-performance data access.
- Bloom Filter Support: Uses Bloom Filters to efficiently detect data existence, reducing access frequency to remote caches.
- Adaptive TTL: Dynamically adjusts TTL based on data access frequency, improving hit rate and cache utilization.
- Prefetching: Supports prefetching, proactively loading data based on usage patterns to further reduce latency.
- Resilient Circuit Breaker and Retry Mechanism: Built-in circuit breakers and retry mechanisms to enhance system stability and fault tolerance.
- Highly Configurable: Offers various configuration options to suit different application needs and scenarios.
- Lightweight Dependencies: Minimal dependencies for easy integration and deployment.
go get -u goflare.io/emberBelow is a simple usage example:
package main
import (
"context"
"fmt"
"log"
"time"
"github.com/redis/go-redis/v9"
"go.uber.org/zap"
"goflare.io/ember"
)
func main() {
// 創建上下文
ctx := context.Background()
// 初始化 Logger
logger, err := zap.NewDevelopment()
if err != nil {
log.Fatalf("Failed to initialize logger: %v", err)
}
defer logger.Sync()
// 配置 Redis 選項
redisOptions := &redis.Options{
Addr: "localhost:6379", // 根據實際情況設置
Password: "", // 無密碼
DB: 0, // 默認數據庫
}
// 初始化 Ember 庫
emberCache, err := ember.New(
ctx,
redisOptions,
ember.WithLogger(logger),
ember.WithMaxLocalSize(2*1024*1024*1024), // 2GB
ember.WithShardCount(16),
ember.WithDefaultExpiration(10*time.Minute),
ember.WithSerialization("json"),
)
if err != nil {
log.Fatalf("Failed to initialize Ember: %v", err)
}
defer func() {
if err := emberCache.Close(); err != nil {
logger.Error("Failed to close Ember", zap.Error(err))
}
}()
// 設置快取項目
key := "example_key"
value := "Hello, Ember!"
ttl := 5 * time.Minute
if err := emberCache.Set(ctx, key, value, ttl); err != nil {
log.Fatalf("Set failed: %v", err)
}
fmt.Printf("Set key '%s' with value '%s' and TTL %v\n", key, value, ttl)
// 獲取快取項目
var retrieved string
found, err := emberCache.Get(ctx, key, &retrieved)
if err != nil {
log.Fatalf("Get failed: %v", err)
}
if found {
fmt.Printf("Retrieved key '%s' with value '%s'\n", key, retrieved)
} else {
fmt.Printf("Key '%s' not found\n", key)
}
// 刪除快取項目
if err := emberCache.Delete(ctx, key); err != nil {
log.Fatalf("Delete failed: %v", err)
}
fmt.Printf("Deleted key '%s'\n", key)
// 清空所有快取項目
if err := emberCache.Clear(ctx); err != nil {
log.Fatalf("Clear failed: %v", err)
}
fmt.Println("Cleared all cache entries")
}Ember offers a rich set of configuration options. Below are the main configuration parameters:
- EnableLocalCache: Enable or disable local caching, default is
true. - MaxLocalSize: Maximum size for the local cache in bytes.
- ShardCount: Number of shards for local cache, recommended to set 2 to 4 times the CPU core count.
- DefaultExpiration: Default data expiration time.
- CleanupInterval: Interval for cleaning up expired items.
- CacheBehaviorConfig: Includes configurations for Prefetching, Adaptive TTL, and Bloom Filter.
- ResilienceConfig: Includes configurations for circuit breakers and retry mechanisms.
- Serialization: Specifies the serialization and deserialization methods for data.
We welcome all forms of contribution! Please see CONTRIBUTING.md for more information.
Ember is distributed under the MIT License. For more details, see LICENSE.