feat(ollama): async rate-limit reset for Ollama Cloud usage windows

Real Ollama Cloud 429s are driven by ollama.com usage windows, not by
response headers that OpenAI/Anthropic/CN rate-limit parsing understands.
Handle them separately: apply a never-shortening temporary cooldown first,
then schedule an async probe that learns the true usage-window reset point
with a controlled, coalesced per-api_key refresh.

- ratelimit_service: route real Ollama Cloud usage accounts to a dedicated
  429 path before the CN-provider branch.
- ollama_cloud_usage: add a background probe coordinator (started in Start,
  waited by Stop) that fans out model-429 probes to one controlled refresh
  per api_key group, reusing the recent group snapshot for coalesced members.
- repository/account_repo: add SetRateLimitedIfUnchanged, a write-back CAS
  counterpart to ClearRateLimitIfObserved, so an async reset never overwrites
  a newer 429, an admin clear, a re-armed generation, or a key/state change
  observed by another writer between read and write.
- wire: inject OllamaCloudUsageService into RateLimitService.

Tests: unit tests for service and repository pass; targeted race tests pass;
build and vet clean; git diff --check clean. The CAS integration test compiles
but is not run (no Docker available).
This commit is contained in:
alfadb
2026-09-13 10:41:47 +08:00
parent bdb42e22f8
commit a1d5968b24
12 changed files with 2311 additions and 5 deletions
+4 -4
View File
@@ -117,10 +117,12 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
openAI403CounterCache := repository.NewOpenAI403CounterCache(redisClient)
geminiTokenCache := repository.NewGeminiTokenCache(redisClient)
compositeTokenCacheInvalidator := service.NewCompositeTokenCacheInvalidator(geminiTokenCache)
rateLimitService := service.ProvideRateLimitService(accountRepository, usageLogRepository, configConfig, geminiQuotaService, tempUnschedCache, timeoutCounterCache, openAI403CounterCache, settingService, compositeTokenCacheInvalidator)
httpUpstream := repository.NewHTTPUpstream(configConfig)
leaderLockCache := repository.NewLeaderLockCache(redisClient)
ollamaCloudUsageService := service.ProvideOllamaCloudUsageService(accountRepository, httpUpstream, settingService, secretEncryptor, configConfig, leaderLockCache, db)
rateLimitService := service.ProvideRateLimitService(accountRepository, usageLogRepository, configConfig, geminiQuotaService, tempUnschedCache, timeoutCounterCache, openAI403CounterCache, settingService, compositeTokenCacheInvalidator, ollamaCloudUsageService)
identityCache := repository.NewIdentityCache(redisClient)
identityService := service.NewIdentityService(identityCache)
httpUpstream := repository.NewHTTPUpstream(configConfig)
timingWheelService, err := service.ProvideTimingWheelService()
if err != nil {
return nil, err
@@ -183,7 +185,6 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
dashboardAggregationRepository := repository.NewDashboardAggregationRepository(db)
dashboardStatsCache := repository.NewDashboardCache(redisClient, configConfig)
dashboardService := service.NewDashboardService(usageLogRepository, dashboardAggregationRepository, dashboardStatsCache, configConfig)
leaderLockCache := repository.NewLeaderLockCache(redisClient)
dashboardAggregationService := service.ProvideDashboardAggregationService(dashboardAggregationRepository, timingWheelService, leaderLockCache, db, configConfig)
dashboardHandler := admin.NewDashboardHandler(dashboardService, dashboardAggregationService)
adminGroupRepository := repository.NewAdminGroupRepository(client, db)
@@ -284,7 +285,6 @@ func initializeApplication(buildInfo handler.BuildInfo) (*Application, error) {
auditLogService := service.ProvideAuditLogService(auditLogRepository, settingService)
auditLogHandler := admin.NewAuditLogHandler(auditLogService, totpService)
upstreamBillingProbeService := service.ProvideUpstreamBillingProbeService(accountRepository, accountTestService, settingService, leaderLockCache, db)
ollamaCloudUsageService := service.ProvideOllamaCloudUsageService(accountRepository, httpUpstream, settingService, secretEncryptor, configConfig, leaderLockCache, db)
adminHandlers := handler.ProvideAdminHandlers(dashboardHandler, adminUserHandler, groupHandler, accountHandler, adminAnnouncementHandler, dataManagementHandler, backupHandler, oAuthHandler, openAIOAuthHandler, geminiOAuthHandler, antigravityOAuthHandler, grokOAuthHandler, cnProviderHandler, proxyHandler, adminRedeemHandler, promoHandler, settingHandler, opsHandler, systemHandler, adminSubscriptionHandler, adminUsageHandler, userAttributeHandler, errorPassthroughHandler, tlsFingerprintProfileHandler, pluginHandler, adminAPIKeyHandler, scheduledTestHandler, channelHandler, channelMonitorHandler, channelMonitorRequestTemplateHandler, contentModerationHandler, promptAdminHandler, paymentHandler, affiliateHandler, complianceHandler, auditLogHandler, upstreamBillingProbeService, ollamaCloudUsageService)
usageRecordWorkerPool := service.NewUsageRecordWorkerPool(configConfig)
userMsgQueueCache := repository.NewUserMsgQueueCache(redisClient)