Commit Graph
39 Commits
Author SHA1 Message Date
William DrakeandClaude Opus 5.5 52c123e6f4 fix(billing): hold inflight reservation until billing lands; price estimate like billing; wire remaining endpoints
Address gaps found in adversarial verification of the in-flight balance
reservation:

- Early release: the reservation is now a ref-counted handle carried on the
  request context. Billing tasks submitted via submit*UsageRecordTask take a
  ref and return it only after RecordUsage has run; with reservation enabled
  the balance cache is deducted synchronously inside the billing task, so the
  reservation is released only after admission can see the deduction. Dropped
  tasks return their ref immediately. The reservation is renewed every ttl/3
  while the handler runs; after the handler returns it is no longer renewed,
  so the post-handler hold is bounded by one TTL.
- Estimate: use the same billing model (channel mapping + BillingModelSource,
  max over requested/mapped when ambiguous) and ModelPricingResolver (group
  ModelPricing incl. wildcards, channel per_request/image/video/token
  overrides), default/group/user rate + peak multiplier, plus image, video,
  audio and search surcharge kinds. Unpriced requests fail open with a
  throttled log; new billing.inflight_reservation.fail_closed_on_unpriced
  (default false) rejects them instead.
- Wire images, embeddings, web/x search, alpha search, grok media generation,
  grok realtime/voice, OpenAI-platform /v1/messages and ResponsesWebSocket
  (session-level reservation with renewal).

Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
2026-09-28 16:14:43 -04:00
Heatherm Huang 9fedf899e2 fix(images): route compatible Gemini image models through API keys 2026-09-12 14:34:31 +08:00
IanShaw 16b15e870d 修复 5888 与 5925 的同号重试语义冲突 2026-08-20 18:16:38 -07:00
IanShaw027 7c53a842e4 合并上游 main,保留 5925 的 Grok 重试上限与 5888 的协议兼容修复
同账号重试采用次数上限加 deadline;畸形 tools 在出站前删除;compaction 422 与结构化错误扫描一并保留。
2026-08-21 08:49:44 +08:00
IanShaw c628b3eea7 让 Grok stream idle 重试上限作用于主路径 2026-08-20 07:07:55 -07:00
IanShaw ad26172b83 完善 Grok 限流冷却与用量兼容 2026-08-20 02:32:43 -07:00
IanShaw c374ff2951 完善 OpenAI 网关切换与运维错误语义 2026-08-19 23:13:40 -07:00
li 02fbcbe3ad fix(ratelimit): 守卫按端点来源门控,并与冷却键对齐模型口径
上一版守卫只看模型类型,不区分请求从哪个端点进来。OAuth 账号的 /v1/images/*
上游同样是 Codex Responses(openai_images_responses.go → handleOpenAIImagesErrorResponse
→ handleOpenAIAccountUpstreamError → HandleUpstreamModelNotFound),所以专用生图
端点也会命中 plan-gated 分支。账号确实不具备生图能力时跳过冷却,会让调度层失去
唯一的刹车:每个请求都完整走一遍号池,对上游形成无上界的 400 放大。

改动:
- 新增 ctxkey.OpenAIImagesEndpoint 与 WithOpenAIImagesEndpoint /
  OpenAIImagesEndpointFromContext,在 handler/openai_images.go 入口置位;
  与 OpenAIImageGenerationIntent 区分——后者在 /v1/responses 带图片模型时也会置位。
- 守卫下移到 modelKey 计算之后,抽成 shouldSkipCodexPlanGatedImageModelCooldown,
  仅在 plan-gated 分支、且非 /v1/images/* 入站时生效。
- 同时判断 requestedModel 与最终 modelKey:冷却键走 account.GetMappedModel,
  账号可以把文本别名映射到 gpt-image-*,只判请求模型会漏掉这种形态。
2026-08-07 20:49:39 +08:00
shaw dec47e8fae fix(profit-control): stop leaking profit policy, close veto livelock, restore passthrough turn pricing
审计修复,逐条如下。

H1 利润策略泄露给所有普通用户
  profit_control_enabled / profit_min_margin / profit_safety_buffer 从
  dto.Group 移到 dto.AdminGroup(后者内嵌前者),赋值相应从
  groupFromServiceBase 移到 GroupFromServiceAdmin;前端 TS 同步从 Group 移到
  AdminGroup。dto.Group 是 GET /api/v1/groups/available 的响应体,该响应本就带
  rate_multiplier,相乘即可反推运营方上游采购成本上限。
  api_contract_test.go 的 /groups/available golden JSON 回滚这三个字段,并把
  fixture 改成非零值(require.JSONEq 是精确比对,缺字段即失败)。
  新增 dto 层边界测试:普通用户 DTO 不含三字段、管理员 DTO 仍含。

M1 利润终检 continue 与 failover 503 退避互动产生活锁
  FailoverState 新增 profitVetoedAccountIDs / profitVetoCount 与
  RecordProfitVeto():加入排除集 + 计数,达 maxProfitVetoAttempts(10) 返回
  FailoverExhausted。HandleSelectionExhausted 的 503 清空分支改为清空后把利润
  否决的账号放回排除集;若排除集已全部由利润否决贡献,清空不会带来任何新候选,
  直接判定耗尽(否则 SwitchCount 永不前进、退避条件永远成立,每 2s 空转一轮)。
  五个 handler 否决点(gateway_handler ×2 / responses / chat_completions /
  gemini_v1beta)改为经 RecordProfitVeto 决策,耗尽时按无可用账号终止。
  回归测试钉死:503 之后持续利润否决必须有限步终止且不 spin;未启用利润控制的
  请求退避语义完全不变。

M2 排队等槽后才终检,延迟可放大到 N × WaitPlan.Timeout
  OpenAI 侧选号循环(自有 failedAccountIDs map,非 FailoverState)新增
  recordOpenAIProfitVeto + handleOpenAIProfitVetoExhausted,共用同一上限语义。
  覆盖 responses / messages-dispatch / chat_completions / alpha_search /
  embeddings / images / grok_media 七处,以及 WS 两处否决分支。

M4 ws_v2 透传 ingress 绕过 per-turn 重定价(选方案 B:最小止血)
  透传 relay 只回调 AfterTurn、没有任何 turn 起始回调,hooks.BeforeTurn 永远
  不触发,而 handler 把 turnPricingAt 初始化成建连时刻 ⇒ 透传连接全部 turn 按
  建连时刻的高峰因子结算,客户端峰前建连保活即可全程谷价——正是本 PR 想堵的
  漏洞。改为 openAIWSTurnPricing 零值起步、只由 BeforeTurn 冻结;透传路径保持
  零值,RecordUsage 回退记录时刻,与引入利润控制前的基线一致。
  未选方案 A(给透传补 turn 起始回调):passthrough_relay.go 是 #5167 刚修过的
  取消传播/close frame 时序敏感区;且 BeforeTurn 还承担 turn>1 的并发槽位抢占,
  接进去等于给透传连接引入 per-turn 抢槽,风险远超本次修复范围。透传仍有建连时
  的准入门,只是没有 turn 级复核,已在两处注释写明。
  测试:service 层钉死透传 ingress 不触发 BeforeTurn(含失败时的复核指引),
  handler 层钉死零值语义与逐 turn 覆盖。

M5 装门读分组走了带账号计数聚合的 GetByID
  SchedulerSnapshotService 新增 GetGroupByIDLite,openai/gateway 两处装门改用
  之。门只需要平台/倍率/利润/高峰字段,且该查询发生在「是否启用利润控制」判定
  之前,未启用的分组同样付代价。两个测试 stub 的 GetByID 改成 panic 守卫。

M6 认证快照注释与真实读取路径相反
  门解析优先取 ctxkey.Group,而它就是本快照物化出来的对象,直连流量走的正是这
  条路。改正注释,与 api_key_repo.go 投影处的说明对齐,避免后人照旧注释删列。

M3 rate_multiplier 为 nil 时利润门 fail-closed(不改行为,加护栏)
  保留 fail-closed。补 repository 层测试钉死账号调度快照的 full/metadata 两份
  payload 都必须保留 RateMultiplier(含 0 值),漏列在 CI 就红。

L1 迁移号注释 191 / 191-192 改为实际的 192/193。
L2 admin group Create 的利润配置预校验改用与 CreateGroup 一致的归一化平台
   (新增 service.NormalizeGroupPlatform,两边共用)。保留预校验而非删除:
   service 层返回的是无类型 error,经 ErrorFrom 会变成 500,删掉会把合法的
   400 降级成 500。
L3 前端利润校验的上界改为判定换算后的小数(后端按小数校验 [0,1)),
   99.999% 会四舍五入进位成 1.0 而被后端 400;i18n en/zh 同步改为 0-99.99。
L4 clampProfitControlThreshold / profitControlOverThreshold 抽为共用函数,
   线上装门/否决点与 profit-preview 不再各自实现,附边界语义测试。
L5 profit-preview 补「默认 D 有账号但最低有效 D 归零」的告警(两档都为 0 由
   既有告警覆盖,不重复)。
2026-08-01 22:39:33 +08:00
Brisbanehuang 20ad5ec506 feat(scheduler): per-group profit control for token account admission
Group pricing (rate multiplier, peak windows, per-user overrides) and
account cost (accounts.rate_multiplier) already live side by side, but
nothing stops the scheduler from handing a request to an account whose
cost multiplier exceeds what the group's pricing can profitably serve.
Add an opt-in per-group profit gate that filters scheduling candidates
by a margin rule, while ordering, scoring, stickiness and breakers keep
working unchanged among qualified accounts.

Admission rule: an account qualifies iff U <= D * (1 - min_margin -
safety_buffer) within a small relative epsilon, where U is
accounts.rate_multiplier (0 is legal; missing/negative/NaN/Inf are
conservatively rejected as invalid) and D is the requester's effective
downstream multiplier (user-group override ?? group default, times the
group peak factor) frozen at the request's pricing instant.

- groups gain profit_control_enabled / profit_min_margin /
  profit_safety_buffer (migration 191); the durable auth-cache
  invalidation trigger additionally watches the profit and pricing
  columns (migration 192) so out-of-band group edits cannot leave
  stale auth snapshots; GetByKeyForAuth explicitly projects the new
  columns and the API-key auth snapshot version is bumped to force a
  refresh of pre-existing snapshots
- request-level pricing instant: token entry points install pricingAt
  into ctx; the profit threshold D and the RecordUsage peak factor
  read the same instant, so one request never changes price mid-flight
  across waits/retries/failover (media and unwired paths keep the
  existing record-time semantics)
- the gate covers token requests on openai, anthropic, gemini, grok
  and antigravity groups: OpenAI-family handlers via
  WithOpenAIRequestPricingContext (responses incl. WS bridge, chat
  completions, messages, embeddings, alpha search), the shared gateway
  via WithGatewayTokenRequestPricing (messages, chat completions,
  responses, gemini model actions); composite groups cannot enable it
  directly; image/video/models/usage/count_tokens stay ungated and an
  explicit image-generation intent suppresses the gate end to end
- post-slot recheck: after a slot is acquired the account is re-read
  via SchedulerSnapshotService.GetAccount (scheduler cache, then DB;
  only when both fail the check fails open with WARN + metric); a
  vetoed account releases its slot and joins the request's exclusion
  set for reselection; sticky bindings are written only after the
  final check passes, and an over-threshold sticky account is skipped,
  not deleted, so it comes back once its rate recovers
- sticky-session cache contract: GatewayCache.GetSessionAccountID now
  returns ErrStickySessionNotFound on a miss (mapped from redis.Nil in
  the repository implementation, mirroring ErrRefreshTokenNotFound) so
  the profit sticky path can distinguish "no binding yet" from a real
  read failure without importing the cache driver in service code
- cross-group re-entry (composite parent -> member group) resolves the
  gate against the member group and clears a stale parent gate instead
  of letting a foreign threshold veto accounts
- per-platform/group activity counters (installs, threshold vetoes,
  invalid-rate vetoes, refresh failures) for observability
- admin UI: profit-control section on the five platforms' group forms
  with percent input, validation and platform-switch reset; group
  create/update/duplicate normalize and validate the config at a
  single choke point
- cmd/profit-preview: offline what-if tool that replays the production
  admission semantics over an exported config/account/override/model
  dump, reports per-model admitted-account counts under the default
  and the worst-case (lowest user override) D, and surfaces probe-sync
  staleness as warnings without affecting admission

Tests: service unit coverage for gate resolution/veto/threshold
epsilon/pricing instant/suppress marker/scheduler filtering and
post-slot recheck (incl. -race on the profit surface), unit-tagged
handler slot-recheck and capability-mapping regressions, sqlmock and
real-PostgreSQL integration regressions for the GetByKeyForAuth
projection and the migration-192 trigger watch list, API contract
update, and frontend specs for the five-platform form helpers.
2026-08-01 22:39:31 +08:00
Wesley Liddick 6d956bdc20 Merge pull request #4801 from SemonCat/feat/persist-session-id
feat(usage): persist client session identifiers
2026-07-25 14:03:32 +08:00
Edison42 1c0cb24c7e feat(usage): persist client session identifiers 2026-07-24 01:22:34 +08:00
404QAQ 4fd9182aff fix(images): log requested quality and size 2026-07-23 23:04:00 +08:00
Heatherm Huang 3a683fff55 Fix composite route alias attribution 2026-07-23 09:20:52 +08:00
Heatherm Huang ebc1028771 Add composite group routing 2026-07-23 09:19:24 +08:00
mt21625457 d11bdb13f5 feat(security-audit): add OpenAI-compatible prompt auditing 2026-07-17 00:39:39 +08:00
王鹏 40b8f04a6a fix(openai): scope transient cooldowns by model 2026-07-15 22:52:35 +08:00
shaw a0593b0bf8 fix(gateway): 客户端断开后 failover 静默终止,不再误报 502 账号耗尽
上游请求经 detachUpstreamContext(WithoutCancel) 有意脱离客户端取消(保
计费),但 failover 循环仍用原始 c.Request.Context() 重新选号:客户端
断开后上游返回 520 等可 failover 错误时,重新选号必然得到 context
canceled,被误判为账号耗尽,记录并返回通用 502。

修复:客户端已断开 ⇒ failover 静默终止。

- 新增 failoverClientGone(c):请求 ctx 已取消时先停 compact 心跳
  (建立 happens-before,对齐其它终结路径),响应未提交则标 499
  (statusClientClosedRequest,与并发槽取消路径同惯例)
- 7 个 OpenAI 内联 failover 循环(Responses/Messages/chat_completions/
  embeddings/images/grok_media/alpha_search)加双 guard:换号前 +
  选号失败分支入口;guard 位于 ReportOpenAIAccountScheduleResult(false)
  之后、RecordOpenAIAccountSwitch/池模式重试之前,账号健康副作用
  (service 层 detached ctx)不受影响
- FailoverState.HandleFailoverError/HandleSelectionExhausted 入口加
  ctx.Err() 检查返回 FailoverCanceled,取消不再改动 failover 状态;
  全部 10 个 FailoverCanceled 分支统一调用 failoverClientGone 归类 499
- 上游 detach 与计费设计不变;真实上游 520 事件仍完整落 ops
  (面板显示码 COALESCE(upstream_status_code,status_code)=520,
  错误率/告警口径不变)

测试:新增 openai_responses_failover_cancel_test.go 复现 issue 场景
(520+取消 ⇒ 不切号、499、无 502 终态)+ 在线客户端对照(正常切换、
耗尽 502);failover_loop_test.go 补入口取消用例并修正取消语义断言。

Fixes #4257
2026-07-15 09:29:05 +08:00
Heatherm Huang b32b815e46 fix(grok): harden OAuth pool recovery 2026-07-14 14:55:30 +08:00
zh239ns 002c0b9fda fix(images): keep non-stream requests alive 2026-07-14 07:39:21 +08:00
DaydreamCodingandClaude Sonnet 5 bdf7ead157 feat(spark-shadow): OpenAI Spark 链接型影子账号
背景:gpt-5.3-codex-spark 使用独立于 codex 全局(5h/7d)的配额窗口(数据源是
/wham/usage 响应体的 codex_bengalfox,而非 codex 全局用的 x-codex-* 响应头),且
只能挂在已完成 OAuth 授权的 OpenAI 账号下复用其登录态,不能作为独立账号单独接入。
为此新增“链接型影子账号”(spark shadow account):影子账号本身不持有任何凭据,
通过 parent_account_id 指向母账号,凭据/token/代理透传自母账号并共享母账号的刷新
周期,仅在配额维度(quota_dimension=spark)和用量窗口上与母账号完全独立调度、互不
连坐。

实现:
- 数据模型:migration 154(+154a)给 accounts 表加 parent_account_id /
  quota_dimension 列 + 4 条约束(维度合法 / parent⟺非 global 维度一致 / 禁自指 /
  FK)+ 2 个 CONCURRENTLY 索引(母账号索引 + 每母账号至多一个影子的唯一索引)。
- 创建:POST /api/v1/admin/accounts/:id/shadow(CreateShadow)—— 一母一影(唯一
  索引兜底并发竞态),继承母账号 proxy/分组/并发/优先级(显式传参可覆盖),默认
  model_mapping 恒等映射到 spark(拒绝非 spark 模型),母账号必须是真实的 OpenAI
  OAuth 账号(非影子)。
- 凭据透传:resolveCredentialAccount 把影子解析回母账号,GetAccessToken / 请求头
  / WS 三条路径统一走此函数;影子自身 Credentials 恒为空(仅允许写 model_mapping),
  凭据写入的汇聚点 persistAccountCredentials 对影子早返 no-op,防止误写。
- 调度:parentHealthyForShadow 只看母账号是否仍是 OpenAI OAuth + 凭据/传输是否
  可用(active、token 未过期、未处于 401/刷新失败/传输故障导致的临时不可调度冷却),
  刻意不看母账号的 global 限流窗口——两条 429 道互不连坐。
- 用量:影子的 codex_5h/7d 走 OpenAIQuotaService.QueryUsage(/wham/usage 的
  codex_bengalfox),与母账号走的 WSv2 探测(/responses 头)完全独立的数据源、
  刷新节流与 staleness 判定。
- 备份:ExportData 显式排除影子账号(影子不持凭据,通用凭据型导入强制
  credentials 非空、无法表达父子链接),按 skipped_shadows 计数提示前端。
- 前端:账号操作菜单新增“创建 Spark 影子”入口,影子行展示回填的母账号信息
  (邮箱 / plan / 隐私模式 / 订阅到期 / chatgpt_account_id),批量操作自动跳过
  影子账号。

说明:migrations 目录用完整文件名(而非纯数字前缀)标识迁移,故本次新增的
154_account_spark_shadow.sql / 154a_..._notx.sql 与已有的
154_add_ops_system_logs_api_key_id.sql 按序号共存,与目录里 145/151 已有的
先例一致。

测试:新增约 20 个测试文件,覆盖 handler(CreateShadow 校验 / 母账号信息回填)、
repository(影子 round-trip / 一母一影唯一索引 / 迁移 schema)、service(凭据
透传三路径 / 调度母健康门 / 用量窗口来源与刷新节流 / CRS 母账号不变量 / 各类
早返与 fail-closed 场景)及前端组件(账号列表 / 操作菜单 / 用量重置)。

验证(镜像 CI;golangci-lint 首次全量分析耗时过长被跳过,其余全部实测):
- gofmt -l:干净
- go build ./... / go vet ./...:通过
- go test ./... -count=1:全绿(全部包 ok,含 internal/service、
  internal/repository、migrations)
- go test -tags integration ./internal/repository/... ./internal/service/...
  (真实 Postgres,testcontainers):全绿,含迁移幂等性
  (TestMigrationsRunner_IsIdempotent_AndSchemaIsUpToDate)与影子相关全部用例
- pnpm lint:check / pnpm typecheck / pnpm build(真实 vite 构建)/
  pnpm vitest run:全绿(124 文件 760 用例)

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
2026-07-01 12:21:45 +08:00
mxyhi 82553c4dca fix(openai): preserve quota platform in usage billing 2026-06-28 11:13:09 +08:00
shaw fcd3bc1272 fix: return 404 model_not_found instead of 503 when no account supports the model 2026-06-26 15:38:06 +08:00
wucm667 da30c59923 fix(openai): fail over image server errors 2026-06-09 13:57:45 +08:00
Wesley Liddick eb24485fc6 Merge pull request #3036 from whatIsNextToTheMoon/fix/openai-response-failed-passthrough
fix(openai): preserve upstream response.failed errors
2026-06-06 09:53:08 +08:00
wucm667 36721d35a8 feat(openai): cool down image rate limits by capability 2026-06-05 18:12:33 +08:00
whatIsNextToTheMoon 5bd3d90434 fix(openai): preserve upstream response.failed errors 2026-06-04 11:17:22 +00:00
name 2caee9d884 refactor(gateway): snapshot usage worker inputs 2026-05-30 20:01:39 +08:00
Wey Gu 2bd3125d0f Preserve usage request context 2026-05-28 22:44:25 +08:00
DaydreamCodingandClaude Opus 4.7 6b39b344d8 feat(quota): 用户 × 平台 USD 配额
为用户在 anthropic/openai/gemini/antigravity 四个平台上提供日/周/月
三个窗口的 USD 配额管控。配额语义:未设置=不限制,0=禁用,>0=美元上限。

两层模型:
- 配置层:系统默认配额,以及 email/linuxdo/oidc/wechat/github/google/
  dingtalk 七个鉴权来源的默认配额,存于 settings,以嵌套 JSON 整体读写
  (系统 1 个 key + 每个来源 1 个 key),整体替换语义。
- 运行时层:user_platform_quota 表按用户记录实际配额,与配置层解耦。

后端:新增 ent schema 与 140_user_platform_quotas.sql 迁移、repository
与 service 端口、计费链路集成、管理端与用户端读写接口。
前端:管理端设置页配额编辑、用户配额管理 Modal、用户 Dashboard 展示、
中英文案。

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
2026-05-26 10:49:20 +08:00
shaw 1e406fed52 fix: optimize OpenAI account cooldown scheduling 2026-05-23 10:18:43 +08:00
gaoren002 888cd8092d fix(openai): surface image moderation errors 2026-05-20 09:19:20 +00:00
name 2eb622f2f6 Remove ops retry replay storage 2026-05-19 19:37:41 +08:00
benjaminandSisyphus 429adbc721 fix: 标记 OpenAI 本地调度容量错误
Ultraworked with [Sisyphus](https://github.com/code-yeongyu/oh-my-openagent)

Co-authored-by: Sisyphus <clio-agent@sisyphuslabs.ai>
2026-05-18 16:52:20 +08:00
shaw fff4a300c6 feat(risk-control): add content moderation audit 2026-05-07 09:14:47 +08:00
2ue 6faa344916 feat: add OpenAI image generation controls 2026-05-05 03:26:54 +08:00
gaoren002 615557ec20 fix(openai): avoid implicit image sticky sessions 2026-04-26 17:09:41 +00:00
james-6-23 dc5d42addc feat(rpm): RPM 限流模块优化
P0:
- rpm_override 嵌入 Auth Cache Snapshot,消除每请求 DB 查询 (snapshot v6→v7)
- 429 RPM 响应返回 Retry-After 头(当前分钟剩余秒数)

P1:
- ClearAll 按钮直连 DELETE API,带 loading 防重复
- 新增 GET /admin/users/:id/rpm-status 管理员 RPM 用量查询端点

优化:
- checkRPM 从级联互斥改为并行取最严,user.rpm_limit 作为全局硬上限始终生效
- Override/Group 变更后自动失效 auth cache
- fail-open 语义不变,Redis 故障不阻塞业务
2026-04-23 16:34:37 +08:00
lucas morgan c548021921 feat(openai): 同步生图 API 支持并接入图片计费调度
- 同步 OpenAI 图片生成与编辑接口
- 接入图片请求解析、账号调度、转发与用量记录
- 接入图片计费与图片用量落库
- 限制 OAuth 生图仅支持无显式模型和尺寸的基础请求
2026-04-22 12:30:08 +08:00