309 lines
12 KiB
Go
309 lines
12 KiB
Go
package handler
|
||||
|
|
|
|||
|
|
import (
|
|||
|
|
"context"
|
|||
|
|
"net/http"
|
|||
|
|
"time"
|
|||
|
|
|
|||
|
|
"github.com/gin-gonic/gin"
|
|||
|
|
|
|||
|
|
"github.com/Wei-Shaw/sub2api/internal/pkg/logger"
|
|||
|
|
"github.com/Wei-Shaw/sub2api/internal/service"
|
|||
|
|
"go.uber.org/zap"
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
// TempUnscheduler 用于 HandleFailoverError 中同账号重试耗尽后的临时封禁。
|
|||
|
|
// GatewayService 隐式实现此接口。
|
|||
|
|
type TempUnscheduler interface {
|
|||
|
|
TempUnscheduleRetryableError(ctx context.Context, accountID int64, failoverErr *service.UpstreamFailoverError)
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// FailoverAction 表示 failover 错误处理后的下一步动作
|
|||
|
|
type FailoverAction int
|
|||
|
|
|
|||
|
|
const (
|
|||
|
|
// FailoverContinue 继续循环(同账号重试或切换账号,调用方统一 continue)
|
|||
|
|
FailoverContinue FailoverAction = iota
|
|||
|
|
// FailoverExhausted 切换次数耗尽(调用方应返回错误响应)
|
|||
|
|
FailoverExhausted
|
|||
|
|
// FailoverCanceled context 已取消(调用方应直接 return)
|
|||
|
|
FailoverCanceled
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
const (
|
|||
|
|
// maxSameAccountRetries 同账号重试次数默认上限(针对 RetryableOnSameAccount 错误)。
|
|||
|
|
// 生产调用方通常传入账号级配置 account.GetPoolModeRetryCount(),该常量仅作兜底/测试默认值。
|
|||
|
|
maxSameAccountRetries = 3
|
|||
|
|
// sameAccountRetryDelay 同账号重试间隔
|
|||
|
|
sameAccountRetryDelay = 500 * time.Millisecond
|
|||
|
|
// maxRequestScopedRetryDelay 限制请求级瞬时错误的指数退避上限,避免高重试配置
|
|||
|
|
// 将单次请求拖入分钟级等待。
|
|||
|
|
maxRequestScopedRetryDelay = 8 * time.Second
|
|||
|
|
// singleAccountBackoffDelay 单账号分组 503 退避重试固定延时。
|
|||
|
|
// Service 层在 SingleAccountRetry 模式下已做充分原地重试(最多 3 次、总等待 30s),
|
|||
|
|
// Handler 层只需短暂间隔后重新进入 Service 层即可。
|
|||
|
|
singleAccountBackoffDelay = 2 * time.Second
|
|||
|
|
// maxProfitVetoAttempts 单次请求内允许的分组利润门终检否决次数上限。
|
|||
|
|
// 利润否决不产生上游请求,因此不会推进 SwitchCount;没有独立上限的话,
|
|||
|
|
// 「选号 → 终检否决 → 重选」在候选池与账号快照短暂不一致时可以空转很久。
|
|||
|
|
// 取值与 maxAccountSwitches 默认值一致:混合定价的大分组仍有充分重选机会,
|
|||
|
|
// 同时把整池越线时的无谓选号开销限制在常数级。
|
|||
|
|
maxProfitVetoAttempts = 10
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
// profitVetoExhaustedMessage 是利润否决次数耗尽时返回给客户端的文案。
|
|||
|
|
// 语义上等同于「无可用账号」:候选账号都不满足分组的利润约束。
|
|||
|
|
const profitVetoExhaustedMessage = "No available accounts: all candidates rejected by group profit control"
|
|||
|
|
|
|||
|
|
func sameAccountRetryDelayFor(failoverErr *service.UpstreamFailoverError, retryCount int) time.Duration {
|
|||
|
|
if failoverErr == nil || !failoverErr.RequestScopedTransient || retryCount <= 1 {
|
|||
|
|
return sameAccountRetryDelay
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
delay := sameAccountRetryDelay
|
|||
|
|
for i := 1; i < retryCount; i++ {
|
|||
|
|
if delay >= maxRequestScopedRetryDelay/2 {
|
|||
|
|
return maxRequestScopedRetryDelay
|
|||
|
|
}
|
|||
|
|
delay *= 2
|
|||
|
|
}
|
|||
|
|
return delay
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// FailoverState 跨循环迭代共享的 failover 状态
|
|||
|
|
type FailoverState struct {
|
|||
|
|
SwitchCount int
|
|||
|
|
MaxSwitches int
|
|||
|
|
FailedAccountIDs map[int64]struct{}
|
|||
|
|
SameAccountRetryCount map[int64]int
|
|||
|
|
LastFailoverErr *service.UpstreamFailoverError
|
|||
|
|
ForceCacheBilling bool
|
|||
|
|
hasBoundSession bool
|
|||
|
|
|
|||
|
|
// profitVetoedAccountIDs 记录被分组利润门终检否决的账号,是 FailedAccountIDs
|
|||
|
|
// 的子集。之所以单独维护:HandleSelectionExhausted 的 503 退避分支会清空
|
|||
|
|
// FailedAccountIDs,而利润否决在同一请求内的判定不会改变(下游倍率 D 已在
|
|||
|
|
// 请求开始冻结),被清空的账号会被立即重选并再次否决,形成没有任何上游请求、
|
|||
|
|
// SwitchCount 也不前进的活锁。清空后必须把它们放回排除集。
|
|||
|
|
profitVetoedAccountIDs map[int64]struct{}
|
|||
|
|
// profitVetoCount 本次请求累计的利润否决次数,用于 maxProfitVetoAttempts 上限。
|
|||
|
|
profitVetoCount int
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// NewFailoverState 创建 failover 状态
|
|||
|
|
func NewFailoverState(maxSwitches int, hasBoundSession bool) *FailoverState {
|
|||
|
|
return &FailoverState{
|
|||
|
|
MaxSwitches: maxSwitches,
|
|||
|
|
FailedAccountIDs: make(map[int64]struct{}),
|
|||
|
|
SameAccountRetryCount: make(map[int64]int),
|
|||
|
|
hasBoundSession: hasBoundSession,
|
|||
|
|
profitVetoedAccountIDs: make(map[int64]struct{}),
|
|||
|
|
}
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// RecordProfitVeto 记录一次分组利润门终检否决:把账号加入排除列表(同时登记到
|
|||
|
|
// 利润否决集,使其不被 503 退避分支清掉)并递增否决计数。
|
|||
|
|
//
|
|||
|
|
// 返回 FailoverContinue 表示调用方可以继续重选下一个账号;返回 FailoverExhausted
|
|||
|
|
// 表示本次请求的利润否决次数已达上限,调用方应按「无可用账号」终止,
|
|||
|
|
// 不得继续 continue。
|
|||
|
|
func (s *FailoverState) RecordProfitVeto(accountID int64) FailoverAction {
|
|||
|
|
s.FailedAccountIDs[accountID] = struct{}{}
|
|||
|
|
if s.profitVetoedAccountIDs == nil {
|
|||
|
|
s.profitVetoedAccountIDs = make(map[int64]struct{})
|
|||
|
|
}
|
|||
|
|
s.profitVetoedAccountIDs[accountID] = struct{}{}
|
|||
|
|
s.profitVetoCount++
|
|||
|
|
if s.profitVetoCount >= maxProfitVetoAttempts {
|
|||
|
|
return FailoverExhausted
|
|||
|
|
}
|
|||
|
|
return FailoverContinue
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// ProfitVetoCount 返回本次请求累计的利润否决次数(供日志使用)。
|
|||
|
|
func (s *FailoverState) ProfitVetoCount() int { return s.profitVetoCount }
|
|||
|
|
|
|||
|
|
// allExclusionsAreProfitVetoed 判断排除列表是否已全部由利润门否决贡献。
|
|||
|
|
// 此时清空 FailedAccountIDs 会被原样恢复,退避重试不会带来任何新候选。
|
|||
|
|
func (s *FailoverState) allExclusionsAreProfitVetoed() bool {
|
|||
|
|
if len(s.profitVetoedAccountIDs) == 0 || len(s.FailedAccountIDs) == 0 {
|
|||
|
|
return false
|
|||
|
|
}
|
|||
|
|
for id := range s.FailedAccountIDs {
|
|||
|
|
if _, ok := s.profitVetoedAccountIDs[id]; !ok {
|
|||
|
|
return false
|
|||
|
|
}
|
|||
|
|
}
|
|||
|
|
return true
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// HandleFailoverError 处理 UpstreamFailoverError,返回下一步动作。
|
|||
|
|
// 包含:缓存计费判断、同账号重试、临时封禁、切换计数、Antigravity 延时。
|
|||
|
|
func (s *FailoverState) HandleFailoverError(
|
|||
|
|
ctx context.Context,
|
|||
|
|
gatewayService TempUnscheduler,
|
|||
|
|
accountID int64,
|
|||
|
|
platform string,
|
|||
|
|
retryLimit int,
|
|||
|
|
failoverErr *service.UpstreamFailoverError,
|
|||
|
|
) FailoverAction {
|
|||
|
|
// 客户端已断开:failover 只会用已取消的 context 重新选号并必然失败,
|
|||
|
|
// 不应再被当成账号耗尽处理(误报 502)。
|
|||
|
|
if ctx != nil && ctx.Err() != nil {
|
|||
|
|
return FailoverCanceled
|
|||
|
|
}
|
|||
|
|
s.LastFailoverErr = failoverErr
|
|||
|
|
if failoverErr == nil || !failoverErr.ShouldRetryNextAccount() {
|
|||
|
|
return FailoverExhausted
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 同账号重试不算切换账号,粘性会话仅在实际切换时强制缓存计费。
|
|||
|
|
sameAccountRetry := failoverErr.RetryableOnSameAccount && s.SameAccountRetryCount[accountID] < retryLimit
|
|||
|
|
if needForceCacheBilling(s.hasBoundSession, failoverErr, sameAccountRetry) {
|
|||
|
|
s.ForceCacheBilling = true
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 同账号重试:对 RetryableOnSameAccount 的临时性错误,先在同一账号上重试。
|
|||
|
|
// 重试次数上限 retryLimit 由调用方传入(账号级 pool_mode_retry_count 配置)。
|
|||
|
|
if failoverErr.RetryableOnSameAccount && s.SameAccountRetryCount[accountID] < retryLimit {
|
|||
|
|
s.SameAccountRetryCount[accountID]++
|
|||
|
|
retryDelay := sameAccountRetryDelayFor(failoverErr, s.SameAccountRetryCount[accountID])
|
|||
|
|
logger.FromContext(ctx).Warn("gateway.failover_same_account_retry",
|
|||
|
|
zap.Int64("account_id", accountID),
|
|||
|
|
zap.Int("upstream_status", failoverErr.StatusCode),
|
|||
|
|
zap.Int("same_account_retry_count", s.SameAccountRetryCount[accountID]),
|
|||
|
|
zap.Int("same_account_retry_max", retryLimit),
|
|||
|
|
zap.Duration("retry_delay", retryDelay),
|
|||
|
|
)
|
|||
|
|
if !sleepWithContext(ctx, retryDelay) {
|
|||
|
|
return FailoverCanceled
|
|||
|
|
}
|
|||
|
|
return FailoverContinue
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 同账号重试用尽,执行临时封禁
|
|||
|
|
if failoverErr.RetryableOnSameAccount {
|
|||
|
|
gatewayService.TempUnscheduleRetryableError(ctx, accountID, failoverErr)
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 加入失败列表
|
|||
|
|
s.FailedAccountIDs[accountID] = struct{}{}
|
|||
|
|
|
|||
|
|
// 检查是否耗尽
|
|||
|
|
if s.SwitchCount >= s.MaxSwitches {
|
|||
|
|
return FailoverExhausted
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// 递增切换计数
|
|||
|
|
s.SwitchCount++
|
|||
|
|
logger.FromContext(ctx).Warn("gateway.failover_switch_account",
|
|||
|
|
zap.Int64("account_id", accountID),
|
|||
|
|
zap.Int("upstream_status", failoverErr.StatusCode),
|
|||
|
|
zap.Int("switch_count", s.SwitchCount),
|
|||
|
|
zap.Int("max_switches", s.MaxSwitches),
|
|||
|
|
)
|
|||
|
|
|
|||
|
|
// Antigravity 平台换号线性递增延时
|
|||
|
|
if platform == service.PlatformAntigravity {
|
|||
|
|
delay := time.Duration(s.SwitchCount-1) * time.Second
|
|||
|
|
if !sleepWithContext(ctx, delay) {
|
|||
|
|
return FailoverCanceled
|
|||
|
|
}
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
return FailoverContinue
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// HandleSelectionExhausted 处理选号失败(所有候选账号都在排除列表中)时的退避重试决策。
|
|||
|
|
// 针对 Antigravity 单账号分组的 503 (MODEL_CAPACITY_EXHAUSTED) 场景:
|
|||
|
|
// 清除排除列表、等待退避后重新选号。
|
|||
|
|
//
|
|||
|
|
// 返回 FailoverContinue 时,调用方应设置 SingleAccountRetry context 并 continue。
|
|||
|
|
// 返回 FailoverExhausted 时,调用方应返回错误响应。
|
|||
|
|
// 返回 FailoverCanceled 时,调用方应直接 return。
|
|||
|
|
func (s *FailoverState) HandleSelectionExhausted(ctx context.Context) FailoverAction {
|
|||
|
|
// 客户端已断开时选号失败是 context canceled 的必然结果,
|
|||
|
|
// 不代表账号耗尽,直接按取消终止。
|
|||
|
|
if ctx.Err() != nil {
|
|||
|
|
return FailoverCanceled
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
if s.LastFailoverErr != nil &&
|
|||
|
|
s.LastFailoverErr.StatusCode == http.StatusServiceUnavailable &&
|
|||
|
|
s.SwitchCount <= s.MaxSwitches {
|
|||
|
|
|
|||
|
|
// 排除列表全由利润门否决贡献时,清空后会被原样恢复:退避重试拿不到
|
|||
|
|
// 任何新候选,而利润否决不推进 SwitchCount,退避条件将永远成立。
|
|||
|
|
// 这里直接判定耗尽,避免每 2s 空转一轮的活锁。
|
|||
|
|
if s.allExclusionsAreProfitVetoed() {
|
|||
|
|
logger.FromContext(ctx).Warn("gateway.failover_selection_exhausted_by_profit_veto",
|
|||
|
|
zap.Int("profit_veto_count", s.profitVetoCount),
|
|||
|
|
zap.Int("excluded_accounts", len(s.FailedAccountIDs)),
|
|||
|
|
)
|
|||
|
|
return FailoverExhausted
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
logger.FromContext(ctx).Warn("gateway.failover_single_account_backoff",
|
|||
|
|
zap.Duration("backoff_delay", singleAccountBackoffDelay),
|
|||
|
|
zap.Int("switch_count", s.SwitchCount),
|
|||
|
|
zap.Int("max_switches", s.MaxSwitches),
|
|||
|
|
)
|
|||
|
|
if !sleepWithContext(ctx, singleAccountBackoffDelay) {
|
|||
|
|
return FailoverCanceled
|
|||
|
|
}
|
|||
|
|
logger.FromContext(ctx).Warn("gateway.failover_single_account_retry",
|
|||
|
|
zap.Int("switch_count", s.SwitchCount),
|
|||
|
|
zap.Int("max_switches", s.MaxSwitches),
|
|||
|
|
)
|
|||
|
|
s.FailedAccountIDs = make(map[int64]struct{})
|
|||
|
|
// 利润门否决的账号不参与退避重试的解除:判定依据(冻结的下游倍率)在
|
|||
|
|
// 同一请求内不变,放它们回池只会被再次否决。
|
|||
|
|
for id := range s.profitVetoedAccountIDs {
|
|||
|
|
s.FailedAccountIDs[id] = struct{}{}
|
|||
|
|
}
|
|||
|
|
return FailoverContinue
|
|||
|
|
}
|
|||
|
|
return FailoverExhausted
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// needForceCacheBilling 判断 failover 时是否需要强制缓存计费。
|
|||
|
|
// 粘性会话实际切换账号、或上游明确标记时,将 input_tokens 转为 cache_read 计费。
|
|||
|
|
func needForceCacheBilling(hasBoundSession bool, failoverErr *service.UpstreamFailoverError, sameAccountRetry bool) bool {
|
|||
|
|
return (hasBoundSession && !sameAccountRetry) || (failoverErr != nil && failoverErr.ForceCacheBilling)
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// failoverClientGone 判断下游客户端是否已断开(请求 context 已取消)。
|
|||
|
|
// 客户端断开后 failover 必须静默终止:用已取消的 context 重新选号只会得到
|
|||
|
|
// context.Canceled,并被误报成账号耗尽(通用 502);上游 detach 的在途请求
|
|||
|
|
// 照常完成计费,但不再为无人接收的响应启动新的上游尝试。
|
|||
|
|
// 响应尚未提交时把状态码标记为 499(client closed request),供访问日志归类。
|
|||
|
|
func failoverClientGone(c *gin.Context) bool {
|
|||
|
|
if c == nil || c.Request == nil || c.Request.Context().Err() == nil {
|
|||
|
|
return false
|
|||
|
|
}
|
|||
|
|
// 先停 compact 心跳(接管 ResponseWriter,建立 happens-before),与
|
|||
|
|
// handleStreamingAwareError/errorResponse 等终结路径对齐,避免心跳
|
|||
|
|
// goroutine 与下面的状态标记并发触碰同一 writer。心跳已提交 200 时
|
|||
|
|
// 状态码已固化,不再标 499。
|
|||
|
|
if service.StopOpenAICompactSSEKeepaliveCommitted(c) {
|
|||
|
|
return true
|
|||
|
|
}
|
|||
|
|
if !c.Writer.Written() {
|
|||
|
|
c.Status(statusClientClosedRequest)
|
|||
|
|
}
|
|||
|
|
return true
|
|||
|
|
}
|
|||
|
|
|
|||
|
|
// sleepWithContext 等待指定时长,返回 false 表示 context 已取消。
|
|||
|
|
func sleepWithContext(ctx context.Context, d time.Duration) bool {
|
|||
|
|
if d <= 0 {
|
|||
|
|
return true
|
|||
|
|
}
|
|||
|
|
select {
|
|||
|
|
case <-ctx.Done():
|
|||
|
|
return false
|
|||
|
|
case <-time.After(d):
|
|||
|
|
return true
|
|||
|
|
}
|
|||
|
|
}
|