feat: add quota limit for API key accounts

- Add configurable spending limit (quota_limit) for apikey-type accounts
- Atomic quota accumulation via PostgreSQL JSONB operations on TotalCost
- Scheduler filters out over-quota accounts with outbox-triggered snapshot refresh
- Display quota usage ($used / $limit) in account capacity column
- Add "Reset Quota" action in account menu to reset usage to zero
- Editing account settings preserves quota_used (no accidental reset)
- Covers all 3 billing paths: Anthropic, Gemini, OpenAI RecordUsage

chore: bump version to 0.1.90.4
This commit is contained in:
erio
2026-03-05 20:54:37 +08:00
parent ae5d9c8bfc
commit 05527b13db
24 changed files with 398 additions and 6 deletions

View File

@@ -1228,6 +1228,10 @@ func (s *GatewayService) SelectAccountWithLoadAwareness(ctx context.Context, gro
modelScopeSkippedIDs = append(modelScopeSkippedIDs, account.ID)
continue
}
// 配额检查
if !s.isAccountSchedulableForQuota(account) {
continue
}
// 窗口费用检查(非粘性会话路径)
if !s.isAccountSchedulableForWindowCost(ctx, account, false) {
filteredWindowCost++
@@ -1260,6 +1264,7 @@ func (s *GatewayService) SelectAccountWithLoadAwareness(ctx context.Context, gro
s.isAccountAllowedForPlatform(stickyAccount, platform, useMixed) &&
(requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, stickyAccount, requestedModel)) &&
s.isAccountSchedulableForModelSelection(ctx, stickyAccount, requestedModel) &&
s.isAccountSchedulableForQuota(stickyAccount) &&
s.isAccountSchedulableForWindowCost(ctx, stickyAccount, true) &&
s.isAccountSchedulableForRPM(ctx, stickyAccount, true) { // 粘性会话窗口费用+RPM 检查
@@ -1416,6 +1421,7 @@ func (s *GatewayService) SelectAccountWithLoadAwareness(ctx context.Context, gro
s.isAccountAllowedForPlatform(account, platform, useMixed) &&
(requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) &&
s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) &&
s.isAccountSchedulableForQuota(account) &&
s.isAccountSchedulableForWindowCost(ctx, account, true) &&
s.isAccountSchedulableForRPM(ctx, account, true) { // 粘性会话窗口费用+RPM 检查
@@ -1480,6 +1486,10 @@ func (s *GatewayService) SelectAccountWithLoadAwareness(ctx context.Context, gro
if !s.isAccountSchedulableForModelSelection(ctx, acc, requestedModel) {
continue
}
// 配额检查
if !s.isAccountSchedulableForQuota(acc) {
continue
}
// 窗口费用检查(非粘性会话路径)
if !s.isAccountSchedulableForWindowCost(ctx, acc, false) {
continue
@@ -2113,6 +2123,15 @@ func (s *GatewayService) withWindowCostPrefetch(ctx context.Context, accounts []
return context.WithValue(ctx, windowCostPrefetchContextKey, costs)
}
// isAccountSchedulableForQuota 检查 API Key 账号是否在配额限制内
// 仅适用于配置了 quota_limit 的 apikey 类型账号
func (s *GatewayService) isAccountSchedulableForQuota(account *Account) bool {
if account.Type != AccountTypeAPIKey {
return true
}
return !account.IsQuotaExceeded()
}
// isAccountSchedulableForWindowCost 检查账号是否可根据窗口费用进行调度
// 仅适用于 Anthropic OAuth/SetupToken 账号
// 返回 true 表示可调度false 表示不可调度
@@ -2590,7 +2609,7 @@ func (s *GatewayService) selectAccountForModelWithPlatform(ctx context.Context,
if clearSticky {
_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), sessionHash)
}
if !clearSticky && s.isAccountInGroup(account, groupID) && account.Platform == platform && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if !clearSticky && s.isAccountInGroup(account, groupID) && account.Platform == platform && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForQuota(account) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if s.debugModelRoutingEnabled() {
logger.LegacyPrintf("service.gateway", "[ModelRoutingDebug] legacy routed sticky hit: group_id=%v model=%s session=%s account=%d", derefGroupID(groupID), requestedModel, shortSessionHash(sessionHash), accountID)
}
@@ -2644,6 +2663,9 @@ func (s *GatewayService) selectAccountForModelWithPlatform(ctx context.Context,
if !s.isAccountSchedulableForModelSelection(ctx, acc, requestedModel) {
continue
}
if !s.isAccountSchedulableForQuota(acc) {
continue
}
if !s.isAccountSchedulableForWindowCost(ctx, acc, false) {
continue
}
@@ -2700,7 +2722,7 @@ func (s *GatewayService) selectAccountForModelWithPlatform(ctx context.Context,
if clearSticky {
_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), sessionHash)
}
if !clearSticky && s.isAccountInGroup(account, groupID) && account.Platform == platform && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if !clearSticky && s.isAccountInGroup(account, groupID) && account.Platform == platform && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForQuota(account) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
return account, nil
}
}
@@ -2743,6 +2765,9 @@ func (s *GatewayService) selectAccountForModelWithPlatform(ctx context.Context,
if !s.isAccountSchedulableForModelSelection(ctx, acc, requestedModel) {
continue
}
if !s.isAccountSchedulableForQuota(acc) {
continue
}
if !s.isAccountSchedulableForWindowCost(ctx, acc, false) {
continue
}
@@ -2818,7 +2843,7 @@ func (s *GatewayService) selectAccountWithMixedScheduling(ctx context.Context, g
if clearSticky {
_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), sessionHash)
}
if !clearSticky && s.isAccountInGroup(account, groupID) && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if !clearSticky && s.isAccountInGroup(account, groupID) && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForQuota(account) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if account.Platform == nativePlatform || (account.Platform == PlatformAntigravity && account.IsMixedSchedulingEnabled()) {
if s.debugModelRoutingEnabled() {
logger.LegacyPrintf("service.gateway", "[ModelRoutingDebug] legacy mixed routed sticky hit: group_id=%v model=%s session=%s account=%d", derefGroupID(groupID), requestedModel, shortSessionHash(sessionHash), accountID)
@@ -2874,6 +2899,9 @@ func (s *GatewayService) selectAccountWithMixedScheduling(ctx context.Context, g
if !s.isAccountSchedulableForModelSelection(ctx, acc, requestedModel) {
continue
}
if !s.isAccountSchedulableForQuota(acc) {
continue
}
if !s.isAccountSchedulableForWindowCost(ctx, acc, false) {
continue
}
@@ -2930,7 +2958,7 @@ func (s *GatewayService) selectAccountWithMixedScheduling(ctx context.Context, g
if clearSticky {
_ = s.cache.DeleteSessionAccountID(ctx, derefGroupID(groupID), sessionHash)
}
if !clearSticky && s.isAccountInGroup(account, groupID) && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if !clearSticky && s.isAccountInGroup(account, groupID) && (requestedModel == "" || s.isModelSupportedByAccountWithContext(ctx, account, requestedModel)) && s.isAccountSchedulableForModelSelection(ctx, account, requestedModel) && s.isAccountSchedulableForQuota(account) && s.isAccountSchedulableForWindowCost(ctx, account, true) && s.isAccountSchedulableForRPM(ctx, account, true) {
if account.Platform == nativePlatform || (account.Platform == PlatformAntigravity && account.IsMixedSchedulingEnabled()) {
return account, nil
}
@@ -2975,6 +3003,9 @@ func (s *GatewayService) selectAccountWithMixedScheduling(ctx context.Context, g
if !s.isAccountSchedulableForModelSelection(ctx, acc, requestedModel) {
continue
}
if !s.isAccountSchedulableForQuota(acc) {
continue
}
if !s.isAccountSchedulableForWindowCost(ctx, acc, false) {
continue
}
@@ -6578,6 +6609,13 @@ func (s *GatewayService) RecordUsage(ctx context.Context, input *RecordUsageInpu
s.billingCacheService.QueueUpdateAPIKeyRateLimitUsage(apiKey.ID, cost.ActualCost)
}
// 更新 API Key 账号配额用量
if shouldBill && cost.TotalCost > 0 && account.Type == AccountTypeAPIKey && account.GetQuotaLimit() > 0 {
if err := s.accountRepo.IncrementQuotaUsed(ctx, account.ID, cost.TotalCost); err != nil {
slog.Error("increment account quota used failed", "account_id", account.ID, "cost", cost.TotalCost, "error", err)
}
}
// Schedule batch update for account last_used_at
s.deferredService.ScheduleLastUsedUpdate(account.ID)
@@ -6775,6 +6813,13 @@ func (s *GatewayService) RecordUsageWithLongContext(ctx context.Context, input *
s.billingCacheService.QueueUpdateAPIKeyRateLimitUsage(apiKey.ID, cost.ActualCost)
}
// 更新 API Key 账号配额用量
if shouldBill && cost.TotalCost > 0 && account.Type == AccountTypeAPIKey && account.GetQuotaLimit() > 0 {
if err := s.accountRepo.IncrementQuotaUsed(ctx, account.ID, cost.TotalCost); err != nil {
slog.Error("increment account quota used failed", "account_id", account.ID, "cost", cost.TotalCost, "error", err)
}
}
// Schedule batch update for account last_used_at
s.deferredService.ScheduleLastUsedUpdate(account.ID)