File size: 4,891 Bytes
6fefda3 |
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 |
package balancer
import (
"context"
"github.com/go-resty/resty/v2"
"jetbrains-ai-proxy/internal/types"
"log"
"sync"
"time"
)
// HealthChecker JWT健康检查器
type HealthChecker struct {
balancer JWTBalancer
client *resty.Client
checkInterval time.Duration
timeout time.Duration
maxRetries int
stopChan chan struct{}
wg sync.WaitGroup
running bool
mutex sync.RWMutex
}
// NewHealthChecker 创建健康检查器
func NewHealthChecker(balancer JWTBalancer) *HealthChecker {
client := resty.New().
SetTimeout(10 * time.Second).
SetHeaders(map[string]string{
"Content-Type": "application/json",
})
return &HealthChecker{
balancer: balancer,
client: client,
checkInterval: 30 * time.Second, // 每30秒检查一次
timeout: 10 * time.Second,
maxRetries: 3,
stopChan: make(chan struct{}),
}
}
// Start 启动健康检查
func (hc *HealthChecker) Start() {
hc.mutex.Lock()
defer hc.mutex.Unlock()
if hc.running {
return
}
hc.running = true
hc.wg.Add(1)
go hc.healthCheckLoop()
log.Println("JWT health checker started")
}
// Stop 停止健康检查
func (hc *HealthChecker) Stop() {
hc.mutex.Lock()
defer hc.mutex.Unlock()
if !hc.running {
return
}
hc.running = false
close(hc.stopChan)
hc.wg.Wait()
log.Println("JWT health checker stopped")
}
// healthCheckLoop 健康检查循环
func (hc *HealthChecker) healthCheckLoop() {
defer hc.wg.Done()
ticker := time.NewTicker(hc.checkInterval)
defer ticker.Stop()
// 启动时立即执行一次检查
hc.performHealthCheck()
for {
select {
case <-ticker.C:
hc.performHealthCheck()
case <-hc.stopChan:
return
}
}
}
// performHealthCheck 执行健康检查
func (hc *HealthChecker) performHealthCheck() {
log.Println("Performing JWT health check...")
// 获取所有tokens进行检查
baseBalancer, ok := hc.balancer.(*BaseBalancer)
if !ok {
log.Println("Warning: Cannot access tokens for health check")
return
}
baseBalancer.mutex.RLock()
tokens := make([]string, 0, len(baseBalancer.tokens))
for token := range baseBalancer.tokens {
tokens = append(tokens, token)
}
baseBalancer.mutex.RUnlock()
// 并发检查所有tokens
var wg sync.WaitGroup
for _, token := range tokens {
wg.Add(1)
go func(t string) {
defer wg.Done()
hc.checkTokenHealth(t)
}(token)
}
wg.Wait()
healthyCount := hc.balancer.GetHealthyTokenCount()
totalCount := hc.balancer.GetTotalTokenCount()
log.Printf("Health check completed: %d/%d tokens healthy", healthyCount, totalCount)
}
// checkTokenHealth 检查单个token的健康状态
func (hc *HealthChecker) checkTokenHealth(token string) {
ctx, cancel := context.WithTimeout(context.Background(), hc.timeout)
defer cancel()
// 创建一个简单的测试请求
testRequest := &types.JetbrainsRequest{
Prompt: types.PROMPT,
Profile: "openai-gpt-4o", // 使用一个通用的profile进行测试
Chat: types.ChatField{
MessageField: []types.MessageField{
{
Type: "user_message",
Content: "test", // 简单的测试消息
},
},
},
}
success := false
for retry := 0; retry < hc.maxRetries; retry++ {
if hc.testTokenRequest(ctx, token, testRequest) {
success = true
break
}
// 重试前等待一小段时间
if retry < hc.maxRetries-1 {
time.Sleep(time.Second)
}
}
if success {
hc.balancer.MarkTokenHealthy(token)
} else {
hc.balancer.MarkTokenUnhealthy(token)
log.Printf("JWT token health check failed: %s...", token[:min(len(token), 10)])
}
}
// testTokenRequest 测试token请求
func (hc *HealthChecker) testTokenRequest(ctx context.Context, token string, req *types.JetbrainsRequest) bool {
resp, err := hc.client.R().
SetContext(ctx).
SetHeader(types.JwtTokenKey, token).
SetBody(req).
Post(types.ChatStreamV7)
if err != nil {
log.Printf("Health check request error for token %s...: %v", token[:min(len(token), 10)], err)
return false
}
// 检查响应状态码
if resp.StatusCode() == 200 {
return true
}
// 401表示token无效,403可能表示配额用完但token有效
if resp.StatusCode() == 403 {
// 配额用完但token有效,仍然标记为健康
return true
}
log.Printf("Health check failed for token %s...: status %d",
token[:min(len(token), 10)], resp.StatusCode())
return false
}
// SetCheckInterval 设置检查间隔
func (hc *HealthChecker) SetCheckInterval(interval time.Duration) {
hc.mutex.Lock()
defer hc.mutex.Unlock()
hc.checkInterval = interval
}
// SetTimeout 设置请求超时
func (hc *HealthChecker) SetTimeout(timeout time.Duration) {
hc.mutex.Lock()
defer hc.mutex.Unlock()
hc.timeout = timeout
}
// SetMaxRetries 设置最大重试次数
func (hc *HealthChecker) SetMaxRetries(retries int) {
hc.mutex.Lock()
defer hc.mutex.Unlock()
hc.maxRetries = retries
}
|