Go学习笔记-Performance优化

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305

package main

import (
    "fmt"
    "log"
    "math/rand"
    "os"
    "runtime"
    "runtime/pprof"
    "sort"
    "strings"
    "time"
)

// 1. CPU性能分析
func cpuProfiling() {
    fmt.Println("=== CPU性能分析 ===")
    
    // 创建CPU profile文件
    cpuFile, err := os.Create("cpu.prof")
    if err != nil {
        log.Fatal("创建CPU profile文件失败:", err)
    }
    defer cpuFile.Close()
    
    // 开始CPU profiling
    if err := pprof.StartCPUProfile(cpuFile); err != nil {
        log.Fatal("启动CPU profiling失败:", err)
    }
    defer pprof.StopCPUProfile()
    
    fmt.Println("开始CPU密集型任务...")
    
    // 执行CPU密集型任务
    start := time.Now()
    result := performCPUIntensiveTask()
    duration := time.Since(start)
    
    fmt.Printf("任务完成，结果: %d，耗时: %v\n", result, duration)
    fmt.Println("CPU profile已保存到 cpu.prof")
}

func performCPUIntensiveTask() int {
    // 生成大量随机数并排序
    const size = 1000000
    data := make([]int, size)
    
    for i := 0; i < size; i++ {
        data[i] = rand.Intn(size)
    }
    
    // 多次排序
    for i := 0; i < 10; i++ {
        sort.Ints(data)
        // 打乱数据
        rand.Shuffle(len(data), func(i, j int) {
            data[i], data[j] = data[j], data[i]
        })
    }
    
    return len(data)
}

// 2. 内存性能分析
func memoryProfiling() {
    fmt.Println("\n=== 内存性能分析 ===")
    
    fmt.Println("开始内存密集型任务...")
    
    // 执行内存密集型任务
    start := time.Now()
    result := performMemoryIntensiveTask()
    duration := time.Since(start)
    
    fmt.Printf("任务完成，结果: %d，耗时: %v\n", result, duration)
    
    // 强制垃圾回收
    runtime.GC()
    
    // 创建内存profile文件
    memFile, err := os.Create("mem.prof")
    if err != nil {
        log.Fatal("创建内存profile文件失败:", err)
    }
    defer memFile.Close()
    
    // 写入内存profile
    if err := pprof.WriteHeapProfile(memFile); err != nil {
        log.Fatal("写入内存profile失败:", err)
    }
    
    fmt.Println("内存profile已保存到 mem.prof")
}

func performMemoryIntensiveTask() int {
    // 创建大量切片和映射
    var slices [][]int
    maps := make(map[int][]string)
    
    for i := 0; i < 1000; i++ {
        // 创建大切片
        slice := make([]int, 10000)
        for j := range slice {
            slice[j] = rand.Intn(1000)
        }
        slices = append(slices, slice)
        
        // 创建字符串切片
        strings := make([]string, 100)
        for j := range strings {
            strings[j] = fmt.Sprintf("string_%d_%d", i, j)
        }
        maps[i] = strings
    }
    
    return len(slices) + len(maps)
}

// 3. 运行时统计信息
func runtimeStats() {
    fmt.Println("\n=== 运行时统计信息 ===")
    
    var m runtime.MemStats
    
    // 获取内存统计信息
    runtime.ReadMemStats(&m)
    
    fmt.Printf("内存统计信息:\n")
    fmt.Printf("  分配的对象数: %d\n", m.Mallocs)
    fmt.Printf("  释放的对象数: %d\n", m.Frees)
    fmt.Printf("  活跃对象数: %d\n", m.Mallocs-m.Frees)
    fmt.Printf("  堆内存大小: %d KB\n", bToKb(m.HeapAlloc))
    fmt.Printf("  堆系统内存: %d KB\n", bToKb(m.HeapSys))
    fmt.Printf("  堆空闲内存: %d KB\n", bToKb(m.HeapIdle))
    fmt.Printf("  堆使用内存: %d KB\n", bToKb(m.HeapInuse))
    fmt.Printf("  栈内存大小: %d KB\n", bToKb(m.StackSys))
    fmt.Printf("  GC次数: %d\n", m.NumGC)
    fmt.Printf("  GC暂停时间: %v\n", time.Duration(m.PauseTotalNs))
    
    // 获取goroutine数量
    fmt.Printf("  Goroutine数量: %d\n", runtime.NumGoroutine())
    fmt.Printf("  CPU核心数: %d\n", runtime.NumCPU())
    fmt.Printf("  Go版本: %s\n", runtime.Version())
}

func bToKb(b uint64) uint64 {
    return b / 1024
}

// 4. 垃圾回收监控
func gcMonitoring() {
    fmt.Println("\n=== 垃圾回收监控 ===")
    
    // 获取GC统计信息
    var stats runtime.MemStats
    runtime.ReadMemStats(&stats)
    
    fmt.Printf("GC统计信息:\n")
    fmt.Printf("  GC次数: %d\n", stats.NumGC)
    fmt.Printf("  强制GC次数: %d\n", stats.NumForcedGC)
    fmt.Printf("  GC总暂停时间: %v\n", time.Duration(stats.PauseTotalNs))
    
    if stats.NumGC > 0 {
        avgPause := time.Duration(stats.PauseTotalNs) / time.Duration(stats.NumGC)
        fmt.Printf("  平均GC暂停时间: %v\n", avgPause)
    }
    
    // 显示最近的GC暂停时间
    fmt.Printf("  最近10次GC暂停时间:\n")
    for i := 0; i < 10 && i < int(stats.NumGC); i++ {
        idx := (int(stats.NumGC) - 1 - i) % len(stats.PauseNs)
        pause := time.Duration(stats.PauseNs[idx])
        fmt.Printf("    GC #%d: %v\n", int(stats.NumGC)-i, pause)
    }
    
    // 手动触发GC
    fmt.Println("\n手动触发GC...")
    before := time.Now()
    runtime.GC()
    gcTime := time.Since(before)
    fmt.Printf("手动GC耗时: %v\n", gcTime)
}

// 5. 性能基准测试
func performanceBenchmark() {
    fmt.Println("\n=== 性能基准测试 ===")
    
    // 测试不同的字符串连接方法
    testStringConcatenation()
    
    // 测试不同的切片操作
    testSliceOperations()
    
    // 测试映射操作
    testMapOperations()
}

func testStringConcatenation() {
    fmt.Println("字符串连接性能测试:")
    
    const iterations = 10000
    strings := make([]string, iterations)
    for i := 0; i < iterations; i++ {
        strings[i] = fmt.Sprintf("string_%d", i)
    }
    
    // 方法1: 使用+操作符
    start := time.Now()
    var result1 string
    for _, s := range strings {
        result1 += s
    }
    time1 := time.Since(start)
    
    // 方法2: 使用strings.Builder
    start = time.Now()
    var builder strings.Builder
    for _, s := range strings {
        builder.WriteString(s)
    }
    result2 := builder.String()
    time2 := time.Since(start)
    
    fmt.Printf("  +操作符: %v (长度: %d)\n", time1, len(result1))
    fmt.Printf("  strings.Builder: %v (长度: %d)\n", time2, len(result2))
    fmt.Printf("  性能提升: %.2fx\n", float64(time1)/float64(time2))
}

func testSliceOperations() {
    fmt.Println("\n切片操作性能测试:")
    
    const size = 1000000
    
    // 方法1: 不预分配容量
    start := time.Now()
    var slice1 []int
    for i := 0; i < size; i++ {
        slice1 = append(slice1, i)
    }
    time1 := time.Since(start)
    
    // 方法2: 预分配容量
    start = time.Now()
    slice2 := make([]int, 0, size)
    for i := 0; i < size; i++ {
        slice2 = append(slice2, i)
    }
    time2 := time.Since(start)
    
    // 方法3: 直接索引赋值
    start = time.Now()
    slice3 := make([]int, size)
    for i := 0; i < size; i++ {
        slice3[i] = i
    }
    time3 := time.Since(start)
    
    fmt.Printf("  不预分配: %v\n", time1)
    fmt.Printf("  预分配容量: %v\n", time2)
    fmt.Printf("  直接赋值: %v\n", time3)
    fmt.Printf("  预分配提升: %.2fx\n", float64(time1)/float64(time2))
    fmt.Printf("  直接赋值提升: %.2fx\n", float64(time1)/float64(time3))
}

func testMapOperations() {
    fmt.Println("\n映射操作性能测试:")
    
    const size = 100000
    
    // 方法1: 不预分配容量
    start := time.Now()
    map1 := make(map[int]string)
    for i := 0; i < size; i++ {
        map1[i] = fmt.Sprintf("value_%d", i)
    }
    time1 := time.Since(start)
    
    // 方法2: 预分配容量
    start = time.Now()
    map2 := make(map[int]string, size)
    for i := 0; i < size; i++ {
        map2[i] = fmt.Sprintf("value_%d", i)
    }
    time2 := time.Since(start)
    
    fmt.Printf("  不预分配: %v\n", time1)
    fmt.Printf("  预分配容量: %v\n", time2)
    fmt.Printf("  性能提升: %.2fx\n", float64(time1)/float64(time2))
}

func main() {
    fmt.Println("Go性能分析和优化示例")
    fmt.Println("========================")
    
    cpuProfiling()
    memoryProfiling()
    runtimeStats()
    gcMonitoring()
    performanceBenchmark()
    
    fmt.Println("\n分析工具使用说明:")
    fmt.Println("1. 查看CPU profile: go tool pprof cpu.prof")
    fmt.Println("2. 查看内存profile: go tool pprof mem.prof")
    fmt.Println("3. 在pprof交互模式中使用 top, list, web 等命令")
}

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336

package main

import (
    "context"
    "fmt"
    "runtime"
    "sync"
    "sync/atomic"
    "time"
)

// 1. 工作池模式
type WorkerPool struct {
    workerCount int
    jobQueue    chan Job
    wg          sync.WaitGroup
}

type Job struct {
    ID   int
    Data interface{}
}

func NewWorkerPool(workerCount, queueSize int) *WorkerPool {
    return &WorkerPool{
        workerCount: workerCount,
        jobQueue:    make(chan Job, queueSize),
    }
}

func (wp *WorkerPool) Start(ctx context.Context) {
    for i := 0; i < wp.workerCount; i++ {
        wp.wg.Add(1)
        go wp.worker(ctx, i)
    }
}

func (wp *WorkerPool) worker(ctx context.Context, id int) {
    defer wp.wg.Done()
    
    for {
        select {
        case job := <-wp.jobQueue:
            wp.processJob(job, id)
        case <-ctx.Done():
            fmt.Printf("Worker %d 停止\n", id)
            return
        }
    }
}

func (wp *WorkerPool) processJob(job Job, workerID int) {
    // 模拟工作处理
    time.Sleep(time.Millisecond * 100)
    fmt.Printf("Worker %d 处理任务 %d\n", workerID, job.ID)
}

func (wp *WorkerPool) Submit(job Job) {
    wp.jobQueue <- job
}

func (wp *WorkerPool) Stop() {
    close(wp.jobQueue)
    wp.wg.Wait()
}

// 2. 工作池性能测试
func workerPoolPerformance() {
    fmt.Println("=== 工作池性能测试 ===")
    
    const jobCount = 1000
    const workerCount = runtime.NumCPU()
    
    fmt.Printf("CPU核心数: %d\n", runtime.NumCPU())
    fmt.Printf("工作者数量: %d\n", workerCount)
    fmt.Printf("任务数量: %d\n", jobCount)
    
    // 创建工作池
    pool := NewWorkerPool(workerCount, 100)
    
    ctx, cancel := context.WithTimeout(context.Background(), 30*time.Second)
    defer cancel()
    
    // 启动工作池
    start := time.Now()
    pool.Start(ctx)
    
    // 提交任务
    go func() {
        for i := 0; i < jobCount; i++ {
            pool.Submit(Job{ID: i, Data: fmt.Sprintf("data_%d", i)})
        }
    }()
    
    // 等待所有任务完成
    time.Sleep(15 * time.Second) // 给足够时间处理所有任务
    cancel()
    pool.Stop()
    
    duration := time.Since(start)
    fmt.Printf("总耗时: %v\n", duration)
    fmt.Printf("平均每个任务: %v\n", duration/jobCount)
}

// 3. 原子操作性能对比
func atomicOperationsPerformance() {
    fmt.Println("\n=== 原子操作性能对比 ===")
    
    const iterations = 10000000
    const goroutines = 10
    
    // 测试1: 使用互斥锁
    var counter1 int64
    var mu sync.Mutex
    var wg1 sync.WaitGroup
    
    start := time.Now()
    for i := 0; i < goroutines; i++ {
        wg1.Add(1)
        go func() {
            defer wg1.Done()
            for j := 0; j < iterations/goroutines; j++ {
                mu.Lock()
                counter1++
                mu.Unlock()
            }
        }()
    }
    wg1.Wait()
    mutexTime := time.Since(start)
    
    // 测试2: 使用原子操作
    var counter2 int64
    var wg2 sync.WaitGroup
    
    start = time.Now()
    for i := 0; i < goroutines; i++ {
        wg2.Add(1)
        go func() {
            defer wg2.Done()
            for j := 0; j < iterations/goroutines; j++ {
                atomic.AddInt64(&counter2, 1)
            }
        }()
    }
    wg2.Wait()
    atomicTime := time.Since(start)
    
    fmt.Printf("互斥锁方式: %v (结果: %d)\n", mutexTime, counter1)
    fmt.Printf("原子操作: %v (结果: %d)\n", atomicTime, counter2)
    fmt.Printf("性能提升: %.2fx\n", float64(mutexTime)/float64(atomicTime))
}

// 4. 通道性能优化
func channelPerformanceOptimization() {
    fmt.Println("\n=== 通道性能优化 ===")
    
    const messageCount = 1000000
    
    // 测试1: 无缓冲通道
    start := time.Now()
    unbufferedChan := make(chan int)
    
    go func() {
        for i := 0; i < messageCount; i++ {
            unbufferedChan <- i
        }
        close(unbufferedChan)
    }()
    
    count1 := 0
    for range unbufferedChan {
        count1++
    }
    unbufferedTime := time.Since(start)
    
    // 测试2: 有缓冲通道
    start = time.Now()
    bufferedChan := make(chan int, 1000)
    
    go func() {
        for i := 0; i < messageCount; i++ {
            bufferedChan <- i
        }
        close(bufferedChan)
    }()
    
    count2 := 0
    for range bufferedChan {
        count2++
    }
    bufferedTime := time.Since(start)
    
    fmt.Printf("无缓冲通道: %v (处理: %d)\n", unbufferedTime, count1)
    fmt.Printf("有缓冲通道: %v (处理: %d)\n", bufferedTime, count2)
    fmt.Printf("性能提升: %.2fx\n", float64(unbufferedTime)/float64(bufferedTime))
}

// 5. 内存池优化
type ObjectPool struct {
    pool sync.Pool
}

func NewObjectPool() *ObjectPool {
    return &ObjectPool{
        pool: sync.Pool{
            New: func() interface{} {
                return make([]byte, 1024) // 1KB缓冲区
            },
        },
    }
}

func (op *ObjectPool) Get() []byte {
    return op.pool.Get().([]byte)
}

func (op *ObjectPool) Put(obj []byte) {
    op.pool.Put(obj)
}

func objectPoolPerformance() {
    fmt.Println("\n=== 对象池性能测试 ===")
    
    const iterations = 1000000
    
    // 测试1: 不使用对象池
    start := time.Now()
    for i := 0; i < iterations; i++ {
        buffer := make([]byte, 1024)
        // 模拟使用缓冲区
        _ = buffer
    }
    noPoolTime := time.Since(start)
    
    // 测试2: 使用对象池
    pool := NewObjectPool()
    start = time.Now()
    for i := 0; i < iterations; i++ {
        buffer := pool.Get()
        // 模拟使用缓冲区
        _ = buffer
        pool.Put(buffer)
    }
    poolTime := time.Since(start)
    
    fmt.Printf("不使用对象池: %v\n", noPoolTime)
    fmt.Printf("使用对象池: %v\n", poolTime)
    fmt.Printf("性能提升: %.2fx\n", float64(noPoolTime)/float64(poolTime))
}

// 6. 并发安全的计数器
type SafeCounter struct {
    mu    sync.RWMutex
    value int64
}

func (c *SafeCounter) Increment() {
    c.mu.Lock()
    c.value++
    c.mu.Unlock()
}

func (c *SafeCounter) Value() int64 {
    c.mu.RLock()
    defer c.mu.RUnlock()
    return c.value
}

type AtomicCounter struct {
    value int64
}

func (c *AtomicCounter) Increment() {
    atomic.AddInt64(&c.value, 1)
}

func (c *AtomicCounter) Value() int64 {
    return atomic.LoadInt64(&c.value)
}

func counterPerformanceComparison() {
    fmt.Println("\n=== 计数器性能对比 ===")
    
    const iterations = 1000000
    const goroutines = 10
    
    // 测试SafeCounter
    safeCounter := &SafeCounter{}
    var wg1 sync.WaitGroup
    
    start := time.Now()
    for i := 0; i < goroutines; i++ {
        wg1.Add(1)
        go func() {
            defer wg1.Done()
            for j := 0; j < iterations/goroutines; j++ {
                safeCounter.Increment()
            }
        }()
    }
    wg1.Wait()
    safeTime := time.Since(start)
    
    // 测试AtomicCounter
    atomicCounter := &AtomicCounter{}
    var wg2 sync.WaitGroup
    
    start = time.Now()
    for i := 0; i < goroutines; i++ {
        wg2.Add(1)
        go func() {
            defer wg2.Done()
            for j := 0; j < iterations/goroutines; j++ {
                atomicCounter.Increment()
            }
        }()
    }
    wg2.Wait()
    atomicTime := time.Since(start)
    
    fmt.Printf("互斥锁计数器: %v (值: %d)\n", safeTime, safeCounter.Value())
    fmt.Printf("原子计数器: %v (值: %d)\n", atomicTime, atomicCounter.Value())
    fmt.Printf("性能提升: %.2fx\n", float64(safeTime)/float64(atomicTime))
}

func main() {
    fmt.Println("Go并发性能优化示例")
    fmt.Println("===================")
    
    workerPoolPerformance()
    atomicOperationsPerformance()
    channelPerformanceOptimization()
    objectPoolPerformance()
    counterPerformanceComparison()
}