Merge pull request #2776 from mt21625457/fix-http2-bug

[codex] 修复 OpenAI/Codex HTTP/2 响应头超时

backend/internal/config/config.go

@@ -680,6 +680,9 @@ type GatewayConfig struct {
 	// 等待上游响应头的超时时间（秒），0表示无超时
 	// 注意：这不影响流式数据传输，只控制等待响应头的时间
 	ResponseHeaderTimeout int `mapstructure:"response_header_timeout"`
+	// OpenAIResponseHeaderTimeout: OpenAI/Codex 上游等待响应头的超时时间（秒），0表示无超时
+	// OpenAI/Codex 请求可能在上游排队较久；默认不使用通用响应头超时截断。
+	OpenAIResponseHeaderTimeout int `mapstructure:"openai_response_header_timeout"`
 	// 请求体最大字节数，用于网关请求体大小限制
 	MaxBodySize int64 `mapstructure:"max_body_size"`
 	// 非流式上游响应体读取上限（字节），用于防止无界读取导致内存放大
@@ -707,6 +710,8 @@ type GatewayConfig struct {
 	OpenAIPassthroughAllowTimeoutHeaders bool `mapstructure:"openai_passthrough_allow_timeout_headers"`
 	// OpenAIWS: OpenAI Responses WebSocket 配置（默认开启，可按需回滚到 HTTP）
 	OpenAIWS GatewayOpenAIWSConfig `mapstructure:"openai_ws"`
+	// OpenAIHTTP2: OpenAI HTTP 上游协议策略（默认启用 HTTP/2，可按代理能力回退 HTTP/1.1）
+	OpenAIHTTP2 GatewayOpenAIHTTP2Config `mapstructure:"openai_http2"`
 	// ImageConcurrency: 图片生成独立并发限制配置（默认关闭）
 	ImageConcurrency ImageConcurrencyConfig `mapstructure:"image_concurrency"`
 
@@ -785,6 +790,21 @@ type GatewayConfig struct {
 	UserMessageQueue UserMessageQueueConfig `mapstructure:"user_message_queue"`
 }
 
+// GatewayOpenAIHTTP2Config OpenAI HTTP 上游协议配置。
+// 默认启用 HTTP/2；在部分代理不兼容时按策略回退 HTTP/1.1。
+type GatewayOpenAIHTTP2Config struct {
+	// Enabled: 是否启用 OpenAI HTTP/2 优先策略
+	Enabled bool `mapstructure:"enabled"`
+	// AllowProxyFallbackToHTTP1: HTTP/HTTPS 代理出现明确 H2 兼容错误时，临时回退 HTTP/1.1
+	AllowProxyFallbackToHTTP1 bool `mapstructure:"allow_proxy_fallback_to_http1"`
+	// FallbackErrorThreshold: 回退窗口内累计多少次兼容错误后触发回退
+	FallbackErrorThreshold int `mapstructure:"fallback_error_threshold"`
+	// FallbackWindowSeconds: 统计兼容错误的时间窗口（秒）
+	FallbackWindowSeconds int `mapstructure:"fallback_window_seconds"`
+	// FallbackTTLSeconds: 触发后回退 HTTP/1.1 的持续时间（秒）
+	FallbackTTLSeconds int `mapstructure:"fallback_ttl_seconds"`
+}
+
 // UserMessageQueueConfig 用户消息串行队列配置
 // 用于 Anthropic OAuth/SetupToken 账号的用户消息串行化发送
 type UserMessageQueueConfig struct {
@@ -1743,6 +1763,7 @@ func setDefaults() {
 
 	// Gateway
 	viper.SetDefault("gateway.response_header_timeout", 600) // 600秒(10分钟)等待上游响应头，LLM高负载时可能排队较久
+	viper.SetDefault("gateway.openai_response_header_timeout", 0)
 	viper.SetDefault("gateway.log_upstream_error_body", true)
 	viper.SetDefault("gateway.log_upstream_error_body_max_bytes", 2048)
 	viper.SetDefault("gateway.inject_beta_for_apikey", false)
@@ -1798,6 +1819,12 @@ func setDefaults() {
 	viper.SetDefault("gateway.openai_ws.scheduler_score_weights.queue", 0.7)
 	viper.SetDefault("gateway.openai_ws.scheduler_score_weights.error_rate", 0.8)
 	viper.SetDefault("gateway.openai_ws.scheduler_score_weights.ttft", 0.5)
+	// OpenAI HTTP upstream protocol strategy
+	viper.SetDefault("gateway.openai_http2.enabled", true)
+	viper.SetDefault("gateway.openai_http2.allow_proxy_fallback_to_http1", true)
+	viper.SetDefault("gateway.openai_http2.fallback_error_threshold", 2)
+	viper.SetDefault("gateway.openai_http2.fallback_window_seconds", 60)
+	viper.SetDefault("gateway.openai_http2.fallback_ttl_seconds", 600)
 	viper.SetDefault("gateway.image_concurrency.enabled", false)
 	viper.SetDefault("gateway.image_concurrency.max_concurrent_requests", 0)
 	viper.SetDefault("gateway.image_concurrency.overflow_mode", ImageConcurrencyOverflowModeReject)
@@ -2365,6 +2392,12 @@ func (c *Config) Validate() error {
 	if c.Gateway.ProxyProbeResponseReadMaxBytes <= 0 {
 		return fmt.Errorf("gateway.proxy_probe_response_read_max_bytes must be positive")
 	}
+	if c.Gateway.ResponseHeaderTimeout < 0 {
+		return fmt.Errorf("gateway.response_header_timeout must be non-negative")
+	}
+	if c.Gateway.OpenAIResponseHeaderTimeout < 0 {
+		return fmt.Errorf("gateway.openai_response_header_timeout must be non-negative")
+	}
 	if strings.TrimSpace(c.Gateway.ConnectionPoolIsolation) != "" {
 		switch c.Gateway.ConnectionPoolIsolation {
 		case ConnectionPoolIsolationProxy, ConnectionPoolIsolationAccount, ConnectionPoolIsolationAccountProxy:
@@ -2539,6 +2572,15 @@ func (c *Config) Validate() error {
 	if c.Gateway.OpenAIWS.StickyPreviousResponseTTLSeconds < 0 {
 		return fmt.Errorf("gateway.openai_ws.sticky_previous_response_ttl_seconds must be non-negative")
 	}
+	if c.Gateway.OpenAIHTTP2.FallbackErrorThreshold < 0 {
+		return fmt.Errorf("gateway.openai_http2.fallback_error_threshold must be non-negative")
+	}
+	if c.Gateway.OpenAIHTTP2.FallbackWindowSeconds < 0 {
+		return fmt.Errorf("gateway.openai_http2.fallback_window_seconds must be non-negative")
+	}
+	if c.Gateway.OpenAIHTTP2.FallbackTTLSeconds < 0 {
+		return fmt.Errorf("gateway.openai_http2.fallback_ttl_seconds must be non-negative")
+	}
 	if c.Gateway.OpenAIWS.SchedulerScoreWeights.Priority < 0 ||
 		c.Gateway.OpenAIWS.SchedulerScoreWeights.Load < 0 ||
 		c.Gateway.OpenAIWS.SchedulerScoreWeights.Queue < 0 ||

backend/internal/config/config_test.go

@@ -163,6 +163,41 @@ func TestLoadDefaultOpenAIWSConfig(t *testing.T) {
 	}
 }
 
+func TestLoadDefaultOpenAIHTTP2Enabled(t *testing.T) {
+	resetViperWithJWTSecret(t)
+
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.True(t, cfg.Gateway.OpenAIHTTP2.Enabled)
+	require.True(t, cfg.Gateway.OpenAIHTTP2.AllowProxyFallbackToHTTP1)
+}
+
+func TestLoadOpenAIHTTP2DisabledFromEnv(t *testing.T) {
+	resetViperWithJWTSecret(t)
+	t.Setenv("GATEWAY_OPENAI_HTTP2_ENABLED", "false")
+
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.False(t, cfg.Gateway.OpenAIHTTP2.Enabled)
+}
+
+func TestLoadDefaultOpenAIResponseHeaderTimeoutUnlimited(t *testing.T) {
+	resetViperWithJWTSecret(t)
+
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.Equal(t, 0, cfg.Gateway.OpenAIResponseHeaderTimeout)
+}
+
+func TestLoadOpenAIResponseHeaderTimeoutFromEnv(t *testing.T) {
+	resetViperWithJWTSecret(t)
+	t.Setenv("GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT", "1800")
+
+	cfg, err := Load()
+	require.NoError(t, err)
+	require.Equal(t, 1800, cfg.Gateway.OpenAIResponseHeaderTimeout)
+}
+
 func TestLoadOpenAIWSStickyTTLCompatibility(t *testing.T) {
 	resetViperWithJWTSecret(t)
 	t.Setenv("GATEWAY_OPENAI_WS_STICKY_RESPONSE_ID_TTL_SECONDS", "0")
@@ -1220,6 +1255,16 @@ func TestValidateConfigErrors(t *testing.T) {
 			mutate:  func(c *Config) { c.Gateway.MaxBodySize = 0 },
 			wantErr: "gateway.max_body_size",
 		},
+		{
+			name:    "gateway response header timeout",
+			mutate:  func(c *Config) { c.Gateway.ResponseHeaderTimeout = -1 },
+			wantErr: "gateway.response_header_timeout",
+		},
+		{
+			name:    "gateway openai response header timeout",
+			mutate:  func(c *Config) { c.Gateway.OpenAIResponseHeaderTimeout = -1 },
+			wantErr: "gateway.openai_response_header_timeout",
+		},
 		{
 			name:    "gateway max idle conns",
 			mutate:  func(c *Config) { c.Gateway.MaxIdleConns = 0 },
@@ -1275,6 +1320,21 @@ func TestValidateConfigErrors(t *testing.T) {
 			mutate:  func(c *Config) { c.Gateway.OpenAIWS.APIKeyMaxConnsFactor = 0 },
 			wantErr: "gateway.openai_ws.apikey_max_conns_factor",
 		},
+		{
+			name:    "gateway openai http2 fallback threshold",
+			mutate:  func(c *Config) { c.Gateway.OpenAIHTTP2.FallbackErrorThreshold = -1 },
+			wantErr: "gateway.openai_http2.fallback_error_threshold",
+		},
+		{
+			name:    "gateway openai http2 fallback window",
+			mutate:  func(c *Config) { c.Gateway.OpenAIHTTP2.FallbackWindowSeconds = -1 },
+			wantErr: "gateway.openai_http2.fallback_window_seconds",
+		},
+		{
+			name:    "gateway openai http2 fallback ttl",
+			mutate:  func(c *Config) { c.Gateway.OpenAIHTTP2.FallbackTTLSeconds = -1 },
+			wantErr: "gateway.openai_http2.fallback_ttl_seconds",
+		},
 		{
 			name:    "gateway stream data interval range",
 			mutate:  func(c *Config) { c.Gateway.StreamDataIntervalTimeout = 5 },

backend/internal/repository/http_upstream.go

@@ -3,6 +3,8 @@ package repository
 import (
 	"compress/flate"
 	"compress/gzip"
+	"context"
+	"crypto/tls"
 	"errors"
 	"fmt"
 	"io"
@@ -49,6 +51,17 @@ const (
 	defaultMaxUpstreamClients = 5000
 	// defaultClientIdleTTLSeconds: 默认客户端空闲回收阈值（15分钟）
 	defaultClientIdleTTLSeconds = 900
+	// OpenAI HTTP/2 代理回退策略默认值
+	defaultOpenAIHTTP2FallbackErrorThreshold = 2
+	defaultOpenAIHTTP2FallbackWindow         = 60 * time.Second
+	defaultOpenAIHTTP2FallbackTTL            = 10 * time.Minute
+)
+
+const (
+	upstreamProtocolModeDefault          = "default"
+	upstreamProtocolModeOpenAIH1         = "openai_h1"
+	upstreamProtocolModeOpenAIH2         = "openai_h2"
+	upstreamProtocolModeOpenAIH1Fallback = "openai_h1_fallback"
 )
 
 var errUpstreamClientLimitReached = errors.New("upstream client cache limit reached")
@@ -63,14 +76,30 @@ type poolSettings struct {
 	responseHeaderTimeout time.Duration // 等待响应头超时时间
 }
 
+type openAIHTTP2Settings struct {
+	enabled                   bool
+	allowProxyFallbackToHTTP1 bool
+	fallbackErrorThreshold    int
+	fallbackWindow            time.Duration
+	fallbackTTL               time.Duration
+}
+
 // upstreamClientEntry 上游客户端缓存条目
 // 记录客户端实例及其元数据，用于连接池管理和淘汰策略
 type upstreamClientEntry struct {
-	client   *http.Client // HTTP 客户端实例
+	client       *http.Client // HTTP 客户端实例
-	proxyKey string       // 代理标识（用于检测代理变更）
+	proxyKey     string       // 代理标识（用于检测代理变更）
-	poolKey  string       // 连接池配置标识（用于检测配置变更）
+	poolKey      string       // 连接池配置标识（用于检测配置变更）
-	lastUsed int64        // 最后使用时间戳（纳秒），用于 LRU 淘汰
+	protocolMode string       // 协议模式（default/openai_h1/openai_h2/openai_h1_fallback）
-	inFlight int64        // 当前进行中的请求数，>0 时不可淘汰
+	lastUsed     int64        // 最后使用时间戳（纳秒），用于 LRU 淘汰
+	inFlight     int64        // 当前进行中的请求数，>0 时不可淘汰
+}
+
+type openAIHTTP2FallbackState struct {
+	mu            sync.Mutex
+	windowStart   time.Time
+	errorCount    int
+	fallbackUntil time.Time
 }
 
 // httpUpstreamService 通用 HTTP 上游服务
@@ -94,6 +123,8 @@ type httpUpstreamService struct {
 	cfg     *config.Config                  // 全局配置
 	mu      sync.RWMutex                    // 保护 clients map 的读写锁
 	clients map[string]*upstreamClientEntry // 客户端缓存池，key 由隔离策略决定
+	// OpenAI 走 HTTP/HTTPS 代理时的 H2->H1 回退状态（key=标准化 proxyKey）
+	openAIHTTP2Fallbacks sync.Map
 }
 
 // NewHTTPUpstream 创建通用 HTTP 上游服务
@@ -131,9 +162,13 @@ func (s *httpUpstreamService) Do(req *http.Request, proxyURL string, accountID i
 	if err := s.validateRequestHost(req); err != nil {
 		return nil, err
 	}
+	profile := service.HTTPUpstreamProfileDefault
+	if req != nil {
+		profile = service.HTTPUpstreamProfileFromContext(req.Context())
+	}
 
 	// 获取或创建对应的客户端，并标记请求占用
-	entry, err := s.acquireClient(proxyURL, accountID, accountConcurrency)
+	entry, err := s.acquireClientWithProfile(proxyURL, accountID, accountConcurrency, profile)
 	if err != nil {
 		return nil, err
 	}
@@ -141,11 +176,13 @@ func (s *httpUpstreamService) Do(req *http.Request, proxyURL string, accountID i
 	// 执行请求
 	resp, err := entry.client.Do(req)
 	if err != nil {
+		s.recordOpenAIHTTP2Failure(profile, entry.protocolMode, entry.proxyKey, err)
 		// 请求失败，立即减少计数
 		atomic.AddInt64(&entry.inFlight, -1)
 		atomic.StoreInt64(&entry.lastUsed, time.Now().UnixNano())
 		return nil, err
 	}
+	s.recordOpenAIHTTP2Success(profile, entry.protocolMode, entry.proxyKey)
 
 	// 如果上游返回了压缩内容，解压后再交给业务层
 	decompressResponseBody(resp)
@@ -168,6 +205,10 @@ func (s *httpUpstreamService) DoWithTLS(req *http.Request, proxyURL string, acco
 	if profile == nil {
 		return s.Do(req, proxyURL, accountID, accountConcurrency)
 	}
+	upstreamProfile := service.HTTPUpstreamProfileDefault
+	if req != nil {
+		upstreamProfile = service.HTTPUpstreamProfileFromContext(req.Context())
+	}
 
 	targetHost := ""
 	if req != nil && req.URL != nil {
@@ -183,7 +224,7 @@ func (s *httpUpstreamService) DoWithTLS(req *http.Request, proxyURL string, acco
 		return nil, err
 	}
 
-	entry, err := s.acquireClientWithTLS(proxyURL, accountID, accountConcurrency, profile)
+	entry, err := s.acquireClientWithTLS(proxyURL, accountID, accountConcurrency, profile, upstreamProfile)
 	if err != nil {
 		slog.Debug("tls_fingerprint_acquire_client_failed", "account_id", accountID, "error", err)
 		return nil, err
@@ -208,21 +249,23 @@ func (s *httpUpstreamService) DoWithTLS(req *http.Request, proxyURL string, acco
 }
 
 // acquireClientWithTLS 获取或创建带 TLS 指纹的客户端
-func (s *httpUpstreamService) acquireClientWithTLS(proxyURL string, accountID int64, accountConcurrency int, profile *tlsfingerprint.Profile) (*upstreamClientEntry, error) {
+func (s *httpUpstreamService) acquireClientWithTLS(proxyURL string, accountID int64, accountConcurrency int, profile *tlsfingerprint.Profile, upstreamProfile service.HTTPUpstreamProfile) (*upstreamClientEntry, error) {
-	return s.getClientEntryWithTLS(proxyURL, accountID, accountConcurrency, profile, true, true)
+	return s.getClientEntryWithTLS(proxyURL, accountID, accountConcurrency, profile, upstreamProfile, true, true)
 }
 
 // getClientEntryWithTLS 获取或创建带 TLS 指纹的客户端条目
 // TLS 指纹客户端使用独立的缓存键，与普通客户端隔离
-func (s *httpUpstreamService) getClientEntryWithTLS(proxyURL string, accountID int64, accountConcurrency int, profile *tlsfingerprint.Profile, markInFlight bool, enforceLimit bool) (*upstreamClientEntry, error) {
+func (s *httpUpstreamService) getClientEntryWithTLS(proxyURL string, accountID int64, accountConcurrency int, profile *tlsfingerprint.Profile, upstreamProfile service.HTTPUpstreamProfile, markInFlight bool, enforceLimit bool) (*upstreamClientEntry, error) {
 	isolation := s.getIsolationMode()
 	proxyKey, parsedProxy, err := normalizeProxyURL(proxyURL)
 	if err != nil {
 		return nil, err
 	}
+	settings := s.resolvePoolSettings(isolation, accountConcurrency)
+	settings = s.applyProfilePoolSettings(settings, upstreamProfile)
 	// TLS 指纹客户端使用独立的缓存键，加 "tls:" 前缀
-	cacheKey := "tls:" + buildCacheKey(isolation, proxyKey, accountID)
+	cacheKey := "tls:" + buildCacheKey(isolation, proxyKey, accountID, upstreamProtocolModeDefault)
-	poolKey := s.buildPoolKey(isolation, accountConcurrency) + ":tls"
+	poolKey := buildPoolKey(settings, upstreamProtocolModeDefault) + ":tls"
 
 	now := time.Now()
 	nowUnix := now.UnixNano()
@@ -273,7 +316,6 @@ func (s *httpUpstreamService) getClientEntryWithTLS(proxyURL string, accountID i
 
 	// 创建带 TLS 指纹的 Transport
 	slog.Debug("tls_fingerprint_creating_new_client", "account_id", accountID, "cache_key", cacheKey, "proxy", proxyKey)
-	settings := s.resolvePoolSettings(isolation, accountConcurrency)
 	transport, err := buildUpstreamTransportWithTLSFingerprint(settings, parsedProxy, profile)
 	if err != nil {
 		s.mu.Unlock()
@@ -339,7 +381,12 @@ func (s *httpUpstreamService) redirectChecker(req *http.Request, via []*http.Req
 // acquireClient 获取或创建客户端，并标记为进行中请求
 // 用于请求路径，避免在获取后被淘汰
 func (s *httpUpstreamService) acquireClient(proxyURL string, accountID int64, accountConcurrency int) (*upstreamClientEntry, error) {
-	return s.getClientEntry(proxyURL, accountID, accountConcurrency, true, true)
+	return s.acquireClientWithProfile(proxyURL, accountID, accountConcurrency, service.HTTPUpstreamProfileDefault)
+}
+
+// acquireClientWithProfile 获取或创建客户端，并按请求 profile 选择协议策略。
+func (s *httpUpstreamService) acquireClientWithProfile(proxyURL string, accountID int64, accountConcurrency int, profile service.HTTPUpstreamProfile) (*upstreamClientEntry, error) {
+	return s.getClientEntry(proxyURL, accountID, accountConcurrency, profile, true, true)
 }
 
 // getOrCreateClient 获取或创建客户端
@@ -358,13 +405,13 @@ func (s *httpUpstreamService) acquireClient(proxyURL string, accountID int64, ac
 //   - account: 按账户隔离，同一账户共享客户端（代理变更时重建）
 //   - account_proxy: 按账户+代理组合隔离，最细粒度
 func (s *httpUpstreamService) getOrCreateClient(proxyURL string, accountID int64, accountConcurrency int) (*upstreamClientEntry, error) {
-	return s.getClientEntry(proxyURL, accountID, accountConcurrency, false, false)
+	return s.getClientEntry(proxyURL, accountID, accountConcurrency, service.HTTPUpstreamProfileDefault, false, false)
 }
 
 // getClientEntry 获取或创建客户端条目
 // markInFlight=true 时会标记进行中请求，用于请求路径防止被淘汰
 // enforceLimit=true 时会限制客户端数量，超限且无法淘汰时返回错误
-func (s *httpUpstreamService) getClientEntry(proxyURL string, accountID int64, accountConcurrency int, markInFlight bool, enforceLimit bool) (*upstreamClientEntry, error) {
+func (s *httpUpstreamService) getClientEntry(proxyURL string, accountID int64, accountConcurrency int, profile service.HTTPUpstreamProfile, markInFlight bool, enforceLimit bool) (*upstreamClientEntry, error) {
 	// 获取隔离模式
 	isolation := s.getIsolationMode()
 	// 标准化代理 URL 并解析
@@ -372,10 +419,14 @@ func (s *httpUpstreamService) getClientEntry(proxyURL string, accountID int64, a
 	if err != nil {
 		return nil, err
 	}
+	// 根据请求 profile（例如 OpenAI）选择协议模式
+	protocolMode := s.resolveProtocolMode(profile, proxyKey, parsedProxy)
+	settings := s.resolvePoolSettings(isolation, accountConcurrency)
+	settings = s.applyProfilePoolSettings(settings, profile)
 	// 构建缓存键（根据隔离策略不同）
-	cacheKey := buildCacheKey(isolation, proxyKey, accountID)
+	cacheKey := buildCacheKey(isolation, proxyKey, accountID, protocolMode)
 	// 构建连接池配置键（用于检测配置变更）
-	poolKey := s.buildPoolKey(isolation, accountConcurrency)
+	poolKey := buildPoolKey(settings, protocolMode)
 
 	now := time.Now()
 	nowUnix := now.UnixNano()
@@ -418,8 +469,7 @@ func (s *httpUpstreamService) getClientEntry(proxyURL string, accountID int64, a
 	}
 
 	// 缓存未命中或需要重建，创建新客户端
-	settings := s.resolvePoolSettings(isolation, accountConcurrency)
+	transport, err := buildUpstreamTransport(settings, parsedProxy, protocolMode)
-	transport, err := buildUpstreamTransport(settings, parsedProxy)
 	if err != nil {
 		s.mu.Unlock()
 		return nil, fmt.Errorf("build transport: %w", err)
@@ -429,9 +479,10 @@ func (s *httpUpstreamService) getClientEntry(proxyURL string, accountID int64, a
 		client.CheckRedirect = s.redirectChecker
 	}
 	entry := &upstreamClientEntry{
-		client:   client,
+		client:       client,
-		proxyKey: proxyKey,
+		proxyKey:     proxyKey,
-		poolKey:  poolKey,
+		poolKey:      poolKey,
+		protocolMode: protocolMode,
 	}
 	atomic.StoreInt64(&entry.lastUsed, nowUnix)
 	if markInFlight {
@@ -615,22 +666,31 @@ func (s *httpUpstreamService) resolvePoolSettings(isolation string, accountConcu
 	return settings
 }
 
-// buildPoolKey 构建连接池配置键
+func (s *httpUpstreamService) applyProfilePoolSettings(settings poolSettings, profile service.HTTPUpstreamProfile) poolSettings {
-// 用于检测配置变更，配置变更时需要重建客户端
+	if profile != service.HTTPUpstreamProfileOpenAI {
-//
+		return settings
-// 参数:
-//   - isolation: 隔离模式
-//   - accountConcurrency: 账户并发限制
-//
-// 返回:
-//   - string: 配置键
-func (s *httpUpstreamService) buildPoolKey(isolation string, accountConcurrency int) string {
-	if isolation == config.ConnectionPoolIsolationAccount || isolation == config.ConnectionPoolIsolationAccountProxy {
-		if accountConcurrency > 0 {
-			return fmt.Sprintf("account:%d", accountConcurrency)
-		}
 	}
-	return "default"
+	settings.responseHeaderTimeout = 0
+	if s != nil && s.cfg != nil && s.cfg.Gateway.OpenAIResponseHeaderTimeout > 0 {
+		settings.responseHeaderTimeout = time.Duration(s.cfg.Gateway.OpenAIResponseHeaderTimeout) * time.Second
+	}
+	return settings
+}
+
+// buildPoolKey 构建连接池配置键，用于检测连接池配置变更。
+func buildPoolKey(settings poolSettings, protocolMode string) string {
+	base := fmt.Sprintf(
+		"idle:%d|idle_host:%d|max:%d|idle_timeout:%s|header_timeout:%s",
+		settings.maxIdleConns,
+		settings.maxIdleConnsPerHost,
+		settings.maxConnsPerHost,
+		settings.idleConnTimeout,
+		settings.responseHeaderTimeout,
+	)
+	if protocolMode == "" || protocolMode == upstreamProtocolModeDefault {
+		return base
+	}
+	return base + "|proto:" + protocolMode
 }
 
 // buildCacheKey 构建客户端缓存键
@@ -648,15 +708,245 @@ func (s *httpUpstreamService) buildPoolKey(isolation string, accountConcurrency
 //   - proxy 模式: "proxy:{proxyKey}"
 //   - account 模式: "account:{accountID}"
 //   - account_proxy 模式: "account:{accountID}|proxy:{proxyKey}"
-func buildCacheKey(isolation, proxyKey string, accountID int64) string {
+func buildCacheKey(isolation, proxyKey string, accountID int64, protocolMode string) string {
+	var base string
 	switch isolation {
 	case config.ConnectionPoolIsolationAccount:
-		return fmt.Sprintf("account:%d", accountID)
+		base = fmt.Sprintf("account:%d", accountID)
 	case config.ConnectionPoolIsolationAccountProxy:
-		return fmt.Sprintf("account:%d|proxy:%s", accountID, proxyKey)
+		base = fmt.Sprintf("account:%d|proxy:%s", accountID, proxyKey)
 	default:
-		return fmt.Sprintf("proxy:%s", proxyKey)
+		base = fmt.Sprintf("proxy:%s", proxyKey)
 	}
+	if protocolMode != "" && protocolMode != upstreamProtocolModeDefault {
+		base += "|proto:" + protocolMode
+	}
+	return base
+}
+
+func (s *httpUpstreamService) resolveOpenAIHTTP2Settings() openAIHTTP2Settings {
+	settings := openAIHTTP2Settings{
+		enabled:                   false,
+		allowProxyFallbackToHTTP1: true,
+		fallbackErrorThreshold:    defaultOpenAIHTTP2FallbackErrorThreshold,
+		fallbackWindow:            defaultOpenAIHTTP2FallbackWindow,
+		fallbackTTL:               defaultOpenAIHTTP2FallbackTTL,
+	}
+	if s == nil || s.cfg == nil {
+		return settings
+	}
+	cfg := s.cfg.Gateway.OpenAIHTTP2
+	settings.enabled = cfg.Enabled
+	settings.allowProxyFallbackToHTTP1 = cfg.AllowProxyFallbackToHTTP1
+	if cfg.FallbackErrorThreshold > 0 {
+		settings.fallbackErrorThreshold = cfg.FallbackErrorThreshold
+	}
+	if cfg.FallbackWindowSeconds > 0 {
+		settings.fallbackWindow = time.Duration(cfg.FallbackWindowSeconds) * time.Second
+	}
+	if cfg.FallbackTTLSeconds > 0 {
+		settings.fallbackTTL = time.Duration(cfg.FallbackTTLSeconds) * time.Second
+	}
+	return settings
+}
+
+func (s *httpUpstreamService) resolveProtocolMode(profile service.HTTPUpstreamProfile, proxyKey string, parsedProxy *url.URL) string {
+	if profile != service.HTTPUpstreamProfileOpenAI {
+		return upstreamProtocolModeDefault
+	}
+	settings := s.resolveOpenAIHTTP2Settings()
+	if !settings.enabled {
+		return upstreamProtocolModeOpenAIH1
+	}
+	if parsedProxy == nil {
+		return upstreamProtocolModeOpenAIH2
+	}
+	scheme := strings.ToLower(parsedProxy.Scheme)
+	if scheme != "http" && scheme != "https" {
+		return upstreamProtocolModeOpenAIH2
+	}
+	if settings.allowProxyFallbackToHTTP1 && s.isOpenAIHTTP2FallbackActive(proxyKey) {
+		return upstreamProtocolModeOpenAIH1Fallback
+	}
+	return upstreamProtocolModeOpenAIH2
+}
+
+func (s *httpUpstreamService) isOpenAIHTTP2FallbackActive(proxyKey string) bool {
+	raw, ok := s.openAIHTTP2Fallbacks.Load(proxyKey)
+	if !ok {
+		return false
+	}
+	state, ok := raw.(*openAIHTTP2FallbackState)
+	if !ok || state == nil {
+		return false
+	}
+	return state.isFallbackActive(time.Now())
+}
+
+func (s *httpUpstreamService) getOrCreateOpenAIHTTP2FallbackState(proxyKey string) *openAIHTTP2FallbackState {
+	state := &openAIHTTP2FallbackState{}
+	actual, _ := s.openAIHTTP2Fallbacks.LoadOrStore(proxyKey, state)
+	cached, ok := actual.(*openAIHTTP2FallbackState)
+	if !ok || cached == nil {
+		return state
+	}
+	return cached
+}
+
+func isHTTPProxyKey(proxyKey string) bool {
+	return strings.HasPrefix(proxyKey, "http://") || strings.HasPrefix(proxyKey, "https://")
+}
+
+func isOpenAIHTTP2CompatibilityError(err error) bool {
+	if err == nil {
+		return false
+	}
+	if isUpstreamTimeoutError(err) {
+		return false
+	}
+	msg := strings.ToLower(err.Error())
+	if msg == "" {
+		return false
+	}
+	markers := []string{
+		"alpn",
+		"no application protocol",
+		"protocol error",
+		"stream error",
+		"goaway",
+		"refused_stream",
+		"frame too large",
+	}
+	for _, marker := range markers {
+		if strings.Contains(msg, marker) {
+			return true
+		}
+	}
+	return false
+}
+
+func isUpstreamTimeoutError(err error) bool {
+	if err == nil {
+		return false
+	}
+	if errors.Is(err, context.DeadlineExceeded) {
+		return true
+	}
+	var netErr net.Error
+	if errors.As(err, &netErr) && netErr.Timeout() {
+		return true
+	}
+	msg := strings.ToLower(err.Error())
+	if msg == "" {
+		return false
+	}
+	timeoutMarkers := []string{
+		"timeout awaiting response headers",
+		"i/o timeout",
+		"context deadline exceeded",
+		"client.timeout exceeded while awaiting headers",
+		"tls handshake timeout",
+	}
+	for _, marker := range timeoutMarkers {
+		if strings.Contains(msg, marker) {
+			return true
+		}
+	}
+	return false
+}
+
+func (s *httpUpstreamService) recordOpenAIHTTP2Failure(profile service.HTTPUpstreamProfile, protocolMode, proxyKey string, err error) {
+	if profile != service.HTTPUpstreamProfileOpenAI || protocolMode != upstreamProtocolModeOpenAIH2 {
+		return
+	}
+	settings := s.resolveOpenAIHTTP2Settings()
+	if !settings.enabled || !settings.allowProxyFallbackToHTTP1 {
+		return
+	}
+	if !isHTTPProxyKey(proxyKey) || !isOpenAIHTTP2CompatibilityError(err) {
+		return
+	}
+	state := s.getOrCreateOpenAIHTTP2FallbackState(proxyKey)
+	activated, until := state.recordFailure(time.Now(), settings.fallbackErrorThreshold, settings.fallbackWindow, settings.fallbackTTL)
+	if activated {
+		slog.Warn("openai_http2_proxy_fallback_activated",
+			"proxy", proxyKey,
+			"fallback_until", until.Format(time.RFC3339))
+	}
+}
+
+func (s *httpUpstreamService) recordOpenAIHTTP2Success(profile service.HTTPUpstreamProfile, protocolMode, proxyKey string) {
+	if profile != service.HTTPUpstreamProfileOpenAI || protocolMode != upstreamProtocolModeOpenAIH2 {
+		return
+	}
+	if !isHTTPProxyKey(proxyKey) {
+		return
+	}
+	raw, ok := s.openAIHTTP2Fallbacks.Load(proxyKey)
+	if !ok {
+		return
+	}
+	state, ok := raw.(*openAIHTTP2FallbackState)
+	if !ok || state == nil {
+		return
+	}
+	state.resetErrorWindow()
+}
+
+func (s *openAIHTTP2FallbackState) isFallbackActive(now time.Time) bool {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+	if s.fallbackUntil.IsZero() {
+		return false
+	}
+	if now.Before(s.fallbackUntil) {
+		return true
+	}
+	s.fallbackUntil = time.Time{}
+	return false
+}
+
+func (s *openAIHTTP2FallbackState) resetErrorWindow() {
+	s.mu.Lock()
+	defer s.mu.Unlock()
+	s.windowStart = time.Time{}
+	s.errorCount = 0
+}
+
+func (s *openAIHTTP2FallbackState) recordFailure(now time.Time, threshold int, window, ttl time.Duration) (bool, time.Time) {
+	if threshold <= 0 {
+		threshold = defaultOpenAIHTTP2FallbackErrorThreshold
+	}
+	if window <= 0 {
+		window = defaultOpenAIHTTP2FallbackWindow
+	}
+	if ttl <= 0 {
+		ttl = defaultOpenAIHTTP2FallbackTTL
+	}
+
+	s.mu.Lock()
+	defer s.mu.Unlock()
+
+	if !s.fallbackUntil.IsZero() && now.Before(s.fallbackUntil) {
+		return false, s.fallbackUntil
+	}
+	if !s.fallbackUntil.IsZero() && !now.Before(s.fallbackUntil) {
+		s.fallbackUntil = time.Time{}
+	}
+
+	if s.windowStart.IsZero() || now.Sub(s.windowStart) > window {
+		s.windowStart = now
+		s.errorCount = 0
+	}
+	s.errorCount++
+	if s.errorCount < threshold {
+		return false, time.Time{}
+	}
+
+	s.fallbackUntil = now.Add(ttl)
+	s.windowStart = time.Time{}
+	s.errorCount = 0
+	return true, s.fallbackUntil
 }
 
 // normalizeProxyURL 标准化代理 URL
@@ -728,7 +1018,7 @@ func defaultPoolSettings(cfg *config.Config) poolSettings {
 		if cfg.Gateway.IdleConnTimeoutSeconds > 0 {
 			idleConnTimeout = time.Duration(cfg.Gateway.IdleConnTimeoutSeconds) * time.Second
 		}
-		if cfg.Gateway.ResponseHeaderTimeout > 0 {
+		if cfg.Gateway.ResponseHeaderTimeout >= 0 {
 			responseHeaderTimeout = time.Duration(cfg.Gateway.ResponseHeaderTimeout) * time.Second
 		}
 	}
@@ -759,7 +1049,7 @@ func defaultPoolSettings(cfg *config.Config) poolSettings {
 //   - MaxConnsPerHost: 每主机最大连接数（达到后新请求等待）
 //   - IdleConnTimeout: 空闲连接超时（超时后关闭）
 //   - ResponseHeaderTimeout: 等待响应头超时（不影响流式传输）
-func buildUpstreamTransport(settings poolSettings, proxyURL *url.URL) (*http.Transport, error) {
+func buildUpstreamTransport(settings poolSettings, proxyURL *url.URL, protocolMode string) (*http.Transport, error) {
 	transport := &http.Transport{
 		MaxIdleConns:          settings.maxIdleConns,
 		MaxIdleConnsPerHost:   settings.maxIdleConnsPerHost,
@@ -767,6 +1057,17 @@ func buildUpstreamTransport(settings poolSettings, proxyURL *url.URL) (*http.Tra
 		IdleConnTimeout:       settings.idleConnTimeout,
 		ResponseHeaderTimeout: settings.responseHeaderTimeout,
 	}
+	switch protocolMode {
+	case upstreamProtocolModeOpenAIH2:
+		transport.ForceAttemptHTTP2 = true
+	case upstreamProtocolModeOpenAIH1:
+		transport.ForceAttemptHTTP2 = false
+		transport.TLSNextProto = make(map[string]func(string, *tls.Conn) http.RoundTripper)
+	case upstreamProtocolModeOpenAIH1Fallback:
+		// 显式禁用 HTTP/2，确保代理不兼容场景回退到 HTTP/1.1。
+		transport.ForceAttemptHTTP2 = false
+		transport.TLSNextProto = make(map[string]func(string, *tls.Conn) http.RoundTripper)
+	}
 	if err := proxyutil.ConfigureTransportProxy(transport, proxyURL); err != nil {
 		return nil, err
 	}

backend/internal/repository/http_upstream_benchmark_test.go

@@ -45,7 +45,7 @@ func BenchmarkHTTPUpstreamProxyClient(b *testing.B) {
 		settings := defaultPoolSettings(cfg)
 		for i := 0; i < b.N; i++ {
 			// 每次迭代都创建新客户端，包含 Transport 分配
-			transport, err := buildUpstreamTransport(settings, parsedProxy)
+			transport, err := buildUpstreamTransport(settings, parsedProxy, upstreamProtocolModeDefault)
 			if err != nil {
 				b.Fatalf("创建 Transport 失败: %v", err)
 			}

backend/internal/repository/http_upstream_test.go

@@ -1,6 +1,7 @@
 package repository
 
 import (
+	"errors"
 	"io"
 	"net/http"
 	"sync/atomic"
@@ -8,6 +9,8 @@ import (
 	"time"
 
 	"github.com/Wei-Shaw/sub2api/internal/config"
+	"github.com/Wei-Shaw/sub2api/internal/pkg/tlsfingerprint"
+	"github.com/Wei-Shaw/sub2api/internal/service"
 	"github.com/stretchr/testify/require"
 	"github.com/stretchr/testify/suite"
 )
@@ -41,12 +44,23 @@ func (s *HTTPUpstreamSuite) newService() *httpUpstreamService {
 }
 
 // TestDefaultResponseHeaderTimeout 测试默认响应头超时配置
-// 验证未配置时使用 300 秒默认值
+// 验证显式 0 会禁用等待响应头超时
 func (s *HTTPUpstreamSuite) TestDefaultResponseHeaderTimeout() {
 	svc := s.newService()
 	entry := mustGetOrCreateClient(s.T(), svc, "", 0, 0)
 	transport, ok := entry.client.Transport.(*http.Transport)
 	require.True(s.T(), ok, "expected *http.Transport")
+	require.Equal(s.T(), time.Duration(0), transport.ResponseHeaderTimeout, "ResponseHeaderTimeout mismatch")
+}
+
+// TestNilConfigResponseHeaderTimeoutFallback 验证 nil 配置使用代码级兜底值。
+func (s *HTTPUpstreamSuite) TestNilConfigResponseHeaderTimeoutFallback() {
+	up := NewHTTPUpstream(nil)
+	svc, ok := up.(*httpUpstreamService)
+	require.True(s.T(), ok, "expected *httpUpstreamService")
+	entry := mustGetOrCreateClient(s.T(), svc, "", 0, 0)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
 	require.Equal(s.T(), 300*time.Second, transport.ResponseHeaderTimeout, "ResponseHeaderTimeout mismatch")
 }
 
@@ -65,10 +79,130 @@ func (s *HTTPUpstreamSuite) TestCustomResponseHeaderTimeout() {
 // 验证解析失败时拒绝回退到直连模式
 func (s *HTTPUpstreamSuite) TestGetOrCreateClient_InvalidURLReturnsError() {
 	svc := s.newService()
-	_, err := svc.getClientEntry("://bad-proxy-url", 1, 1, false, false)
+	_, err := svc.getClientEntry("://bad-proxy-url", 1, 1, service.HTTPUpstreamProfileDefault, false, false)
 	require.Error(s.T(), err, "expected error for invalid proxy URL")
 }
 
+func (s *HTTPUpstreamSuite) TestOpenAIProfileDefaultsToHTTP2AndNoHeaderTimeout() {
+	s.cfg.Gateway = config.GatewayConfig{
+		ResponseHeaderTimeout: 600,
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{
+			Enabled:                   true,
+			AllowProxyFallbackToHTTP1: true,
+		},
+	}
+	svc := s.newService()
+	entry, err := svc.getClientEntry("", 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.Equal(s.T(), time.Duration(0), transport.ResponseHeaderTimeout, "OpenAI profile should not inherit generic header timeout")
+	require.True(s.T(), transport.ForceAttemptHTTP2, "OpenAI profile should prefer HTTP/2")
+	require.Equal(s.T(), upstreamProtocolModeOpenAIH2, entry.protocolMode)
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIProfileCustomHeaderTimeout() {
+	s.cfg.Gateway = config.GatewayConfig{
+		ResponseHeaderTimeout:       600,
+		OpenAIResponseHeaderTimeout: 1800,
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{
+			Enabled: true,
+		},
+	}
+	svc := s.newService()
+	entry, err := svc.getClientEntry("", 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.Equal(s.T(), 1800*time.Second, transport.ResponseHeaderTimeout)
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIProfileTLSFingerprintDoesNotInheritGenericHeaderTimeout() {
+	s.cfg.Gateway = config.GatewayConfig{
+		ResponseHeaderTimeout: 600,
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{
+			Enabled: true,
+		},
+	}
+	svc := s.newService()
+	entry, err := svc.getClientEntryWithTLS("", 1, 1, &tlsfingerprint.Profile{Name: "test"}, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.Equal(s.T(), time.Duration(0), transport.ResponseHeaderTimeout, "OpenAI TLS path should not inherit generic header timeout")
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIProfileHTTP2DisabledUsesHTTP1Transport() {
+	s.cfg.Gateway = config.GatewayConfig{
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{Enabled: false},
+	}
+	svc := s.newService()
+	entry, err := svc.getClientEntry("", 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.False(s.T(), transport.ForceAttemptHTTP2, "OpenAI HTTP/2 disabled should not force H2")
+	require.NotNil(s.T(), transport.TLSNextProto, "HTTP/1 mode should disable automatic H2 negotiation")
+	require.Equal(s.T(), upstreamProtocolModeOpenAIH1, entry.protocolMode)
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIHeaderTimeoutChangeRebuildsClient() {
+	s.cfg.Gateway = config.GatewayConfig{
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{Enabled: true},
+	}
+	svc := s.newService()
+	entry1, err := svc.getClientEntry("", 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+
+	s.cfg.Gateway.OpenAIResponseHeaderTimeout = 1800
+	entry2, err := svc.getClientEntry("", 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	require.NotSame(s.T(), entry1, entry2, "OpenAI header timeout changes must rebuild cached client")
+	transport, ok := entry2.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.Equal(s.T(), 1800*time.Second, transport.ResponseHeaderTimeout)
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIHTTP2TimeoutDoesNotActivateProxyFallback() {
+	s.cfg.Gateway = config.GatewayConfig{
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{
+			Enabled:                   true,
+			AllowProxyFallbackToHTTP1: true,
+			FallbackErrorThreshold:    1,
+			FallbackWindowSeconds:     60,
+			FallbackTTLSeconds:        600,
+		},
+	}
+	svc := s.newService()
+	proxyURL := "http://proxy.local:8080"
+	svc.recordOpenAIHTTP2Failure(service.HTTPUpstreamProfileOpenAI, upstreamProtocolModeOpenAIH2, proxyURL, errors.New("http2: timeout awaiting response headers"))
+	require.False(s.T(), svc.isOpenAIHTTP2FallbackActive(proxyURL), "header timeout should not be treated as H2 compatibility failure")
+}
+
+func (s *HTTPUpstreamSuite) TestOpenAIHTTP2ProxyCompatibilityErrorActivatesFallback() {
+	s.cfg.Gateway = config.GatewayConfig{
+		OpenAIHTTP2: config.GatewayOpenAIHTTP2Config{
+			Enabled:                   true,
+			AllowProxyFallbackToHTTP1: true,
+			FallbackErrorThreshold:    1,
+			FallbackWindowSeconds:     60,
+			FallbackTTLSeconds:        600,
+		},
+	}
+	svc := s.newService()
+	proxyURL := "http://proxy.local:8080"
+	svc.recordOpenAIHTTP2Failure(service.HTTPUpstreamProfileOpenAI, upstreamProtocolModeOpenAIH2, proxyURL, errors.New("http2: protocol error"))
+	require.True(s.T(), svc.isOpenAIHTTP2FallbackActive(proxyURL))
+
+	entry, err := svc.getClientEntry(proxyURL, 1, 1, service.HTTPUpstreamProfileOpenAI, false, false)
+	require.NoError(s.T(), err)
+	transport, ok := entry.client.Transport.(*http.Transport)
+	require.True(s.T(), ok, "expected *http.Transport")
+	require.False(s.T(), transport.ForceAttemptHTTP2)
+	require.NotNil(s.T(), transport.TLSNextProto)
+	require.Equal(s.T(), upstreamProtocolModeOpenAIH1Fallback, entry.protocolMode)
+}
+
 // TestNormalizeProxyURL_Canonicalizes 测试代理 URL 规范化
 // 验证等价地址能够映射到同一缓存键
 func (s *HTTPUpstreamSuite) TestNormalizeProxyURL_Canonicalizes() {

backend/internal/service/account_test_service.go

@@ -580,6 +580,7 @@ func (s *AccountTestService) testOpenAIAccountConnection(c *gin.Context, account
 	if err != nil {
 		return s.sendErrorAndEnd(c, "Failed to create request")
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 
 	// Set common headers
 	req.Header.Set("Content-Type", "application/json")
@@ -659,6 +660,7 @@ func (s *AccountTestService) testOpenAIChatCompletionsConnection(
 	if err != nil {
 		return s.sendErrorAndEnd(c, "Failed to create Chat Completions request")
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 	req.Header.Set("Content-Type", "application/json")
 	req.Header.Set("Accept", "text/event-stream")
 	req.Header.Set("Authorization", "Bearer "+authToken)
@@ -739,6 +741,7 @@ func (s *AccountTestService) testOpenAICompactConnection(c *gin.Context, account
 	if err != nil {
 		return s.sendErrorAndEnd(c, "Failed to create request")
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 
 	req.Header.Set("Content-Type", "application/json")
 	req.Header.Set("Accept", "application/json")
@@ -1505,6 +1508,7 @@ func (s *AccountTestService) testOpenAIImageAPIKey(c *gin.Context, ctx context.C
 	if err != nil {
 		return s.sendErrorAndEnd(c, "Failed to create request")
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 	req.Header.Set("Content-Type", "application/json")
 	req.Header.Set("Authorization", "Bearer "+authToken)
 
@@ -1593,6 +1597,7 @@ func (s *AccountTestService) testOpenAIImageOAuth(c *gin.Context, ctx context.Co
 	if err != nil {
 		return s.sendErrorAndEnd(c, "Failed to create request")
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 	req.Host = "chatgpt.com"
 	req.Header.Set("Authorization", "Bearer "+authToken)
 	req.Header.Set("Content-Type", "application/json")

backend/internal/service/account_test_service_openai_compact_test.go

@@ -57,6 +57,7 @@ func TestAccountTestService_TestAccountConnection_OpenAICompactOAuthSuccessPersi
 	require.Equal(t, "application/json", upstream.lastReq.Header.Get("Accept"))
 	require.Equal(t, codexCLIVersion, upstream.lastReq.Header.Get("Version"))
 	require.NotEmpty(t, upstream.lastReq.Header.Get("Session_Id"))
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Equal(t, codexCLIUserAgent, upstream.lastReq.Header.Get("User-Agent"))
 	require.Equal(t, "chatgpt-acc", upstream.lastReq.Header.Get("chatgpt-account-id"))
 	require.Equal(t, "gpt-5.4", gjson.GetBytes(upstream.lastBody, "model").String())

backend/internal/service/account_test_service_openai_image_test.go

@@ -45,6 +45,8 @@ func TestAccountTestService_OpenAIImageOAuthHandlesOutputItemDoneFallback(t *tes
 
 	err := svc.testOpenAIImageOAuth(c, context.Background(), account, "gpt-image-2", "draw a cat")
 	require.NoError(t, err)
+	require.NotNil(t, upstream.lastReq)
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Contains(t, rec.Body.String(), "Calling Codex /responses image tool")
 	require.Contains(t, rec.Body.String(), "data:image/png;base64,aGVsbG8=")
 	require.Contains(t, rec.Body.String(), "\"success\":true")
@@ -83,6 +85,7 @@ func TestAccountTestService_OpenAIImageAPIKeyUsesConfiguredV1BaseURL(t *testing.
 	err := svc.testOpenAIImageAPIKey(c, context.Background(), account, "gpt-image-2", "draw a cat")
 	require.NoError(t, err)
 	require.NotNil(t, upstream.lastReq)
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Equal(t, "https://image-upstream.example/v1/images/generations", upstream.lastReq.URL.String())
 	require.Equal(t, "Bearer test-api-key", upstream.lastReq.Header.Get("Authorization"))
 	require.Contains(t, rec.Body.String(), "data:image/png;base64,aGVsbG8=")

backend/internal/service/account_test_service_openai_test.go

@@ -129,6 +129,8 @@ func TestAccountTestService_OpenAISuccessPersistsSnapshotFromHeaders(t *testing.
 
 	err := svc.testOpenAIAccountConnection(ctx, account, "gpt-5.4", "", "")
 	require.NoError(t, err)
+	require.Len(t, upstream.requests, 1)
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.requests[0].Context()))
 	require.NotEmpty(t, repo.updatedExtra)
 	require.Equal(t, 42.0, repo.updatedExtra["codex_5h_used_percent"])
 	require.Equal(t, 88.0, repo.updatedExtra["codex_7d_used_percent"])
@@ -372,6 +374,7 @@ func TestAccountTestService_OpenAIAPIKeyResponsesUnsupportedUsesChatCompletionsP
 	err := svc.testOpenAIAccountConnection(ctx, account, "gpt-5.4", "hello", "")
 	require.NoError(t, err)
 	require.NotNil(t, upstream.lastReq)
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Equal(t, "https://compat-upstream.example/v1/chat/completions", upstream.lastReq.URL.String())
 	require.Equal(t, "Bearer sk-test", upstream.lastReq.Header.Get("Authorization"))
 	require.Equal(t, "text/event-stream", upstream.lastReq.Header.Get("Accept"))

backend/internal/service/http_upstream_profile.go

@@ -0,0 +1,42 @@
+package service
+
+import "context"
+
+// HTTPUpstreamProfile marks HTTP upstream requests that need provider-specific
+// transport policy.
+type HTTPUpstreamProfile string
+
+const (
+	HTTPUpstreamProfileDefault HTTPUpstreamProfile = ""
+	HTTPUpstreamProfileOpenAI  HTTPUpstreamProfile = "openai"
+)
+
+type httpUpstreamProfileContextKey struct{}
+
+// WithHTTPUpstreamProfile injects an upstream transport profile into ctx.
+func WithHTTPUpstreamProfile(ctx context.Context, profile HTTPUpstreamProfile) context.Context {
+	if ctx == nil {
+		ctx = context.Background()
+	}
+	if profile == HTTPUpstreamProfileDefault {
+		return ctx
+	}
+	return context.WithValue(ctx, httpUpstreamProfileContextKey{}, profile)
+}
+
+// HTTPUpstreamProfileFromContext resolves the upstream transport profile from ctx.
+func HTTPUpstreamProfileFromContext(ctx context.Context) HTTPUpstreamProfile {
+	if ctx == nil {
+		return HTTPUpstreamProfileDefault
+	}
+	profile, ok := ctx.Value(httpUpstreamProfileContextKey{}).(HTTPUpstreamProfile)
+	if !ok {
+		return HTTPUpstreamProfileDefault
+	}
+	switch profile {
+	case HTTPUpstreamProfileOpenAI:
+		return profile
+	default:
+		return HTTPUpstreamProfileDefault
+	}
+}

backend/internal/service/http_upstream_profile_test.go

@@ -0,0 +1,21 @@
+package service
+
+import (
+	"context"
+	"testing"
+)
+
+func TestWithHTTPUpstreamProfile_DefaultKeepsContext(t *testing.T) {
+	ctx := context.Background()
+	got := WithHTTPUpstreamProfile(ctx, HTTPUpstreamProfileDefault)
+	if got != ctx {
+		t.Fatal("default profile should not wrap context")
+	}
+}
+
+func TestWithHTTPUpstreamProfile_OpenAI(t *testing.T) {
+	ctx := WithHTTPUpstreamProfile(context.TODO(), HTTPUpstreamProfileOpenAI)
+	if profile := HTTPUpstreamProfileFromContext(ctx); profile != HTTPUpstreamProfileOpenAI {
+		t.Fatalf("expected profile %q, got %q", HTTPUpstreamProfileOpenAI, profile)
+	}
+}

backend/internal/service/openai_apikey_responses_probe.go

@@ -95,6 +95,7 @@ func (s *AccountTestService) ProbeOpenAIAPIKeyResponsesSupport(ctx context.Conte
 		logger.LegacyPrintf("service.openai_probe", "probe_build_request_failed: account_id=%d err=%v", accountID, err)
 		return
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 	req.Header.Set("Content-Type", "application/json")
 	req.Header.Set("Authorization", "Bearer "+apiKey)
 	req.Header.Set("Accept", "application/json")

backend/internal/service/openai_gateway_chat_completions_raw.go

@@ -135,6 +135,7 @@ func (s *OpenAIGatewayService) forwardAsRawChatCompletions(
 	if err != nil {
 		return nil, fmt.Errorf("build upstream request: %w", err)
 	}
+	upstreamReq = upstreamReq.WithContext(WithHTTPUpstreamProfile(upstreamReq.Context(), HTTPUpstreamProfileOpenAI))
 	upstreamReq.Header.Set("Content-Type", "application/json")
 	upstreamReq.Header.Set("Authorization", "Bearer "+apiKey)
 	if clientStream {

backend/internal/service/openai_gateway_chat_completions_raw_test.go

@@ -116,6 +116,7 @@ func TestForwardAsRawChatCompletions_ForcesStreamUsageUpstreamAndPassesUsageDown
 	require.Equal(t, 3, result.Usage.CacheReadInputTokens)
 	require.NotNil(t, upstream.lastReq)
 	require.NoError(t, upstream.lastReq.Context().Err())
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.True(t, gjson.GetBytes(upstream.lastBody, "stream_options.include_usage").Bool())
 	require.Contains(t, rec.Body.String(), `"usage"`)
 	require.Contains(t, rec.Body.String(), "data: [DONE]")

backend/internal/service/openai_gateway_responses_chat_fallback.go

@@ -116,6 +116,7 @@ func (s *OpenAIGatewayService) forwardResponsesViaRawChatCompletions(
 	if err != nil {
 		return nil, fmt.Errorf("build upstream request: %w", err)
 	}
+	upstreamReq = upstreamReq.WithContext(WithHTTPUpstreamProfile(upstreamReq.Context(), HTTPUpstreamProfileOpenAI))
 	upstreamReq.Header.Set("Content-Type", "application/json")
 	upstreamReq.Header.Set("Authorization", "Bearer "+apiKey)
 	if clientStream {

backend/internal/service/openai_gateway_responses_chat_fallback_test.go

@@ -42,6 +42,7 @@ func TestForwardResponses_ForceChatCompletionsRoutesNonStreamingToChatCompletion
 	require.NoError(t, err)
 	require.NotNil(t, result)
 	require.Equal(t, "http://upstream.example/v1/chat/completions", upstream.lastReq.URL.String())
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Equal(t, "hello", gjson.GetBytes(upstream.lastBody, "messages.0.content").String())
 	require.False(t, gjson.GetBytes(upstream.lastBody, "input").Exists())
 	require.Equal(t, "response", gjson.Get(rec.Body.String(), "object").String())

backend/internal/service/openai_gateway_service.go

@@ -3229,6 +3229,7 @@ func (s *OpenAIGatewayService) buildUpstreamRequestOpenAIPassthrough(
 	if err != nil {
 		return nil, err
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 
 	// 透传客户端请求头（安全白名单）。
 	allowTimeoutHeaders := s.isOpenAIPassthroughTimeoutHeadersAllowed()
@@ -3951,6 +3952,7 @@ func (s *OpenAIGatewayService) buildUpstreamRequest(ctx context.Context, c *gin.
 	if err != nil {
 		return nil, err
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 
 	// Set authentication header
 	req.Header.Set("authorization", "Bearer "+token)

backend/internal/service/openai_gateway_service_test.go

@@ -1841,6 +1841,7 @@ func TestOpenAIBuildUpstreamRequestOpenAIPassthroughPreservesCompactPath(t *test
 	require.Equal(t, "application/json", req.Header.Get("Accept"))
 	require.Equal(t, codexCLIVersion, req.Header.Get("Version"))
 	require.NotEmpty(t, req.Header.Get("Session_Id"))
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(req.Context()))
 }
 
 func TestOpenAIBuildUpstreamRequestCompactForcesJSONAcceptForOAuth(t *testing.T) {
@@ -1861,6 +1862,7 @@ func TestOpenAIBuildUpstreamRequestCompactForcesJSONAcceptForOAuth(t *testing.T)
 	require.Equal(t, "application/json", req.Header.Get("Accept"))
 	require.Equal(t, codexCLIVersion, req.Header.Get("Version"))
 	require.NotEmpty(t, req.Header.Get("Session_Id"))
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(req.Context()))
 }
 
 func TestOpenAIBuildUpstreamRequestOAuthMessagesBridgeUsesSessionOnly(t *testing.T) {

backend/internal/service/openai_images.go

@@ -743,6 +743,7 @@ func (s *OpenAIGatewayService) buildOpenAIImagesRequest(
 	if err != nil {
 		return nil, err
 	}
+	req = req.WithContext(WithHTTPUpstreamProfile(req.Context(), HTTPUpstreamProfileOpenAI))
 	req.Header.Set("Authorization", "Bearer "+token)
 	for key, values := range c.Request.Header {
 		if !openaiPassthroughAllowedHeaders[strings.ToLower(key)] {

backend/internal/service/openai_images_test.go

@@ -528,6 +528,7 @@ func TestOpenAIGatewayServiceForwardImages_OAuthPassesNAndReturnsAllImages(t *te
 	require.NotNil(t, upstream.lastReq)
 	require.Equal(t, chatgptCodexURL, upstream.lastReq.URL.String())
 	require.Equal(t, "chatgpt.com", upstream.lastReq.Host)
+	require.Equal(t, HTTPUpstreamProfileOpenAI, HTTPUpstreamProfileFromContext(upstream.lastReq.Context()))
 	require.Equal(t, "application/json", upstream.lastReq.Header.Get("Content-Type"))
 	require.Equal(t, "text/event-stream", upstream.lastReq.Header.Get("Accept"))
 	require.Equal(t, "acct-123", upstream.lastReq.Header.Get("chatgpt-account-id"))

deploy/.env.example

@@ -251,6 +251,14 @@ RATE_LIMIT_OVERLOAD_COOLDOWN_MINUTES=10
 #
 # 默认：false
 GATEWAY_FORCE_CODEX_CLI=false
+# OpenAI/Codex 等待上游响应头超时（秒）；0 表示不使用本地响应头超时截断。
+GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT=0
+# OpenAI HTTP 上游默认启用 HTTP/2；如需紧急回滚可设为 false。
+GATEWAY_OPENAI_HTTP2_ENABLED=true
+GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1=true
+GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD=2
+GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS=60
+GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS=600
 # 上游连接池：每主机最大连接数（默认 1024；流式/HTTP1.1 场景可调大，如 2400/4096）
 GATEWAY_MAX_CONNS_PER_HOST=2048
 # 上游连接池：最大空闲连接总数（默认 2560；账号/代理隔离 + 高并发场景可调大）

deploy/config.example.yaml

@@ -149,6 +149,9 @@ gateway:
   # Timeout for waiting upstream response headers (seconds)
   # 等待上游响应头超时时间（秒）
   response_header_timeout: 600
+  # OpenAI/Codex upstream response header timeout (seconds, 0=disabled)
+  # OpenAI/Codex 等待上游响应头超时时间（秒，0=禁用本地响应头超时）
+  openai_response_header_timeout: 0
   # Max request body size in bytes (default: 256MB)
   # 请求体最大字节数（默认 256MB）
   max_body_size: 268435456
@@ -317,6 +320,14 @@ gateway:
       queue: 0.7
       error_rate: 0.8
       ttft: 0.5
+  # OpenAI HTTP upstream protocol strategy.
+  # OpenAI HTTP 上游协议策略（默认 HTTP/2；代理明确不兼容时可临时回退 HTTP/1.1）。
+  openai_http2:
+    enabled: true
+    allow_proxy_fallback_to_http1: true
+    fallback_error_threshold: 2
+    fallback_window_seconds: 60
+    fallback_ttl_seconds: 600
   # HTTP upstream connection pool settings (HTTP/2 + multi-proxy scenario defaults)
   # HTTP 上游连接池配置（HTTP/2 + 多代理场景默认值）
   # Max idle connections across all hosts

deploy/docker-compose.dev.yml

@@ -40,6 +40,13 @@ services:
       - JWT_SECRET=${JWT_SECRET:-}
       - TOTP_ENCRYPTION_KEY=${TOTP_ENCRYPTION_KEY:-}
       - TZ=${TZ:-Asia/Shanghai}
+      # OpenAI HTTP upstream protocol/timeout
+      - GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT=${GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT:-0}
+      - GATEWAY_OPENAI_HTTP2_ENABLED=${GATEWAY_OPENAI_HTTP2_ENABLED:-true}
+      - GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1=${GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1:-true}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD=${GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD:-2}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS:-60}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS:-600}
       - GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT=${GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT:-900}
       - GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL=${GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL:-10}
       - GATEWAY_IMAGE_CONCURRENCY_ENABLED=${GATEWAY_IMAGE_CONCURRENCY_ENABLED:-false}

deploy/docker-compose.local.yml

@@ -151,6 +151,13 @@ services:
       # =======================================================================
       # Image Generation Stream & Concurrency
       # =======================================================================
+      # OpenAI HTTP upstream protocol/timeout
+      - GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT=${GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT:-0}
+      - GATEWAY_OPENAI_HTTP2_ENABLED=${GATEWAY_OPENAI_HTTP2_ENABLED:-true}
+      - GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1=${GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1:-true}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD=${GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD:-2}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS:-60}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS:-600}
       - GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT=${GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT:-900}
       - GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL=${GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL:-10}
       - GATEWAY_IMAGE_CONCURRENCY_ENABLED=${GATEWAY_IMAGE_CONCURRENCY_ENABLED:-false}

deploy/docker-compose.standalone.yml

@@ -98,6 +98,13 @@ services:
       # =======================================================================
       # Image Generation Stream & Concurrency
       # =======================================================================
+      # OpenAI HTTP upstream protocol/timeout
+      - GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT=${GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT:-0}
+      - GATEWAY_OPENAI_HTTP2_ENABLED=${GATEWAY_OPENAI_HTTP2_ENABLED:-true}
+      - GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1=${GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1:-true}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD=${GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD:-2}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS:-60}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS:-600}
       - GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT=${GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT:-900}
       - GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL=${GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL:-10}
       - GATEWAY_IMAGE_CONCURRENCY_ENABLED=${GATEWAY_IMAGE_CONCURRENCY_ENABLED:-false}

deploy/docker-compose.yml

@@ -147,6 +147,13 @@ services:
       # =======================================================================
       # Image Generation Stream & Concurrency
       # =======================================================================
+      # OpenAI HTTP upstream protocol/timeout
+      - GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT=${GATEWAY_OPENAI_RESPONSE_HEADER_TIMEOUT:-0}
+      - GATEWAY_OPENAI_HTTP2_ENABLED=${GATEWAY_OPENAI_HTTP2_ENABLED:-true}
+      - GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1=${GATEWAY_OPENAI_HTTP2_ALLOW_PROXY_FALLBACK_TO_HTTP1:-true}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD=${GATEWAY_OPENAI_HTTP2_FALLBACK_ERROR_THRESHOLD:-2}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_WINDOW_SECONDS:-60}
+      - GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS=${GATEWAY_OPENAI_HTTP2_FALLBACK_TTL_SECONDS:-600}
       - GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT=${GATEWAY_IMAGE_STREAM_DATA_INTERVAL_TIMEOUT:-900}
       - GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL=${GATEWAY_IMAGE_STREAM_KEEPALIVE_INTERVAL:-10}
       - GATEWAY_IMAGE_CONCURRENCY_ENABLED=${GATEWAY_IMAGE_CONCURRENCY_ENABLED:-false}