Commit Graph

1009 Commits

Author SHA1 Message Date
Edric Li
a54b81cf74 perf: 错误处理性能优化
- MatchRule 延迟/限制 body ToLower,先用 statusCode 短路,只在需要关键词匹配时转换且限制 8KB
- 预计算规则的小写关键词/平台和 error code set,消除运行时重复 ToLower 和线性扫描
- MODEL_CAPACITY_EXHAUSTED 全局去重,避免并发请求重复重试同一模型
- 503 重试 body 读取限制从 2MB 降至 8KB
- time.After 替换为 time.NewTimer,防止 context 取消时 timer 泄漏
2026-02-10 21:40:31 +08:00
Edric Li
2d4236f76e fix: 修复错误透传规则 skip_monitoring 未生效的问题
- ops_error_logger: status < 400 分支增加 OpsSkipPassthroughKey 检查
- ops_upstream_context: 新增 checkSkipMonitoringForUpstreamEvent,中间重试/故障转移事件也能触发跳过标记
- gateway_handler/openai_gateway_handler/gemini_v1beta_handler: handleFailoverExhausted 匹配规则后设置 OpsSkipPassthroughKey
- antigravity_gateway_service: writeMappedClaudeError 增加 applyErrorPassthroughRule 调用
2026-02-10 20:56:01 +08:00
Edric Li
7d0a30fa8f merge: sync upstream main (antigravity single-account 503 retry)
合并上游新增的 Antigravity 单账号 503 退避重试机制,
解决与本地 MODEL_CAPACITY_EXHAUSTED 逻辑的冲突,两者共存。
2026-02-10 12:00:21 +08:00
Edric Li
d95e04fd1f feat: 错误透传规则支持 skip_monitoring 跳过运维监控记录
在每条错误透传规则上新增 skip_monitoring 选项,开启后匹配该规则的错误
不会被记录到 ops_error_logs,减少监控噪音。默认关闭,不影响现有规则。
2026-02-10 11:42:39 +08:00
shaw
5dd83d3cf2 fix: 移除特定system以适配新版cc客户端缓存失效的bug 2026-02-10 10:28:34 +08:00
Wesley Liddick
14e1aac9b5 Merge pull request #533 from GuangYiDing/feat/antigravity-single-account-503-retry
feat: Antigravity 单账号分组 503 退避重试机制
2026-02-10 09:59:48 +08:00
Edric Li
6114f69cca feat: MODEL_CAPACITY_EXHAUSTED 使用固定1s间隔重试60次,不切换账号
MODEL_CAPACITY_EXHAUSTED (503) 表示模型容量不足,所有账号共享同一容量池,
切换账号无意义。改为固定1s间隔重试最多60次,重试耗尽后直接返回上游错误。

- 新增 antigravityModelCapacityRetryMaxAttempts=60 和 antigravityModelCapacityRetryWait=1s
- shouldTriggerAntigravitySmartRetry 新增 isModelCapacityExhausted 返回值
- handleSmartRetry 对 MODEL_CAPACITY_EXHAUSTED 使用独立重试策略
- handleModelRateLimit 对 MODEL_CAPACITY_EXHAUSTED 仅标记 Handled,不设限流
- 重试耗尽后不设置模型限流、不清除粘性会话、不切换账号
2026-02-10 02:03:06 +08:00
Edric Li
d6c2921f2b feat: same-account retry before failover for transient errors
For retryable transient errors (Google 400 "invalid project resource name"
and empty stream responses), retry on the same account up to 2 times
(with 500ms delay) before switching to another account.

- Add RetryableOnSameAccount field to UpstreamFailoverError
- Add same-account retry loop in both Gemini and Claude/OpenAI handler paths
- Move temp-unschedule from service layer to handler layer (only after
  all same-account retries exhausted)
- Reduce temp-unschedule cooldown from 30 minutes to 1 minute
2026-02-10 00:53:54 +08:00
Edric Li
61c73287dc feat: failover and temp-unschedule on empty stream response
- Empty stream responses now return UpstreamFailoverError instead of
  plain 502, triggering automatic account switching (up to 10 retries)
- Add tempUnscheduleEmptyResponse: accounts returning empty responses
  are temp-unscheduled for 30 minutes
- Apply to both Claude and Gemini non-streaming paths
- Align googleConfigErrorCooldown from 60m to 30m for consistency
2026-02-09 23:25:30 +08:00
Edric Li
89905ec43d feat: failover and temp-unschedule on Google "Invalid project resource name" 400
Google 后端间歇性返回 400 "Invalid project resource name" 错误,
此前该错误直接透传给客户端且不触发账号切换,导致请求失败。

- 在 Antigravity 和 Gemini 两个平台的所有转发路径中,
  精确匹配该错误消息后触发 failover 自动换号重试
- 命中后将账号临时封禁 1 小时,避免反复调度到同一故障账号
- 提取共享函数 isGoogleProjectConfigError / tempUnscheduleGoogleConfigError
  消除跨 Service 的代码重复
2026-02-09 22:48:32 +08:00
Rose Ding
e4bc35151f test: 添加单账号 503 退避重试机制的单元测试
覆盖 Service 层和 Handler 层的所有新增逻辑:
- isSingleAccountRetry context 标记检查
- handleSmartRetry 中 503 + SingleAccountRetry 分支
- handleSingleAccountRetryInPlace 原地重试逻辑
- antigravityRetryLoop 预检查跳过限流
- sleepAntigravitySingleAccountBackoff 固定延迟退避
- 端到端集成场景验证

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
2026-02-09 22:06:06 +08:00
Wesley Liddick
56da498b7e Merge pull request #532 from touwaeriol/fix/clear-model-rate-limits
fix: support clearing model-level rate limits from action menu and temp-unsched reset
2026-02-09 20:52:44 +08:00
erio
4a84ca9a02 fix: support clearing model-level rate limits from action menu and temp-unsched reset 2026-02-09 20:37:30 +08:00
erio
a70d37a676 fix: Gemini error policy check should precede retry logic 2026-02-09 19:55:17 +08:00
erio
6892e84ad2 fix: skip rate limiting when custom error codes don't match upstream status
Add ShouldHandleErrorCode guard at the entry of handleGeminiUpstreamError
and AntigravityGatewayService.handleUpstreamError so that accounts with
custom error codes (e.g. [599]) are not rate-limited when the upstream
returns a non-matching status (e.g. 429).
2026-02-09 19:55:05 +08:00
erio
73f455745c feat: ErrorPolicySkipped returns 500 instead of upstream status code
When custom error codes are enabled and the upstream error code is NOT
in the configured list, return HTTP 500 to the client instead of
transparently forwarding the original status code.

Also adds integration test TestCustomErrorCode599 verifying that 429,
500, 503, 401, 403 all return 500 without triggering SetRateLimited
or SetError.
2026-02-09 19:54:54 +08:00
Rose Ding
021abfca18 fix: 单账号分组首次 503 不设模型限流标记,避免后续请求雪崩
单账号 antigravity 分组收到 503 (MODEL_CAPACITY_EXHAUSTED) 时,
原逻辑会设置 ~29s 模型限流标记。由于只有一个账号无法切换,
后续所有新请求在预检查时命中限流 → 几毫秒内直接返回 503,
导致约 30 秒的雪崩窗口。

修复:在 Handler 入口处检查分组是否只有单个 antigravity 账号,
如果是则提前设置 SingleAccountRetry context 标记,让 Service 层
首次 503 就走原地重试逻辑(不设限流标记),避免污染后续请求。
2026-02-09 17:25:36 +08:00
Rose Ding
f6cfab9901 feat: 添加 Antigravity 单账号 503 退避重试机制
当分组内只有一个可用账号且上游返回 503 (MODEL_CAPACITY_EXHAUSTED) 时,
不再设置模型限流+切换账号(因为切换回来还是同一个账号),而是在 Service 层
原地等待+重试,避免双重等待问题。

主要变更:
- Handler 层:检测单账号 503 场景,清除排除列表并设置 SingleAccountRetry 标记
- Service 层:新增 handleSingleAccountRetryInPlace 原地重试逻辑
- Service 层:预检查跳过单账号模式下的限流检查
- 新增 ctxkey.SingleAccountRetry 上下文标记
2026-02-09 14:26:01 +08:00
shaw
51572b5da0 chore: update version 2026-02-09 12:00:03 +08:00
QTom
04cedce9a1 test: 为 stubAccountRepo 添加 ListCRSAccountIDs 方法实现 2026-02-09 11:40:37 +08:00
QTom
5e0d789440 feat(admin): 新增 CRS 同步预览和账号选择功能
- 后端新增 PreviewFromCRS 接口,允许用户先预览 CRS 中的账号
- 后端支持在同步时选择特定账号,不选中的账号将被跳过
- 前端重构 SyncFromCrsModal 为三步向导:输入凭据 → 预览账号 → 执行同步
- 改进表单无障碍性:添加 for/id 关联和 required 属性
- 修复 Back 按钮返回时的状态清理
- 新增 buildSelectedSet 和 shouldCreateAccount 的单元测试
- 完整的向后兼容性:旧客户端不发送 selected_account_ids 时行为不变
2026-02-09 10:39:09 +08:00
erio
9a479d1b55 fix: resolve CI failures from scope removal refactor
- Fix gofmt alignment in ops_realtime_models.go
- Remove SetAntigravityQuotaScopeLimit mock from api_contract_test.go
- Add UpdateSortOrders mock to mockGroupRepoForGateway
2026-02-09 08:27:14 +08:00
erio
fc095bf054 refactor: replace scope-level rate limiting with model-level rate limiting
Merge functional changes from develop branch:
- Remove AntigravityQuotaScope system (claude/gemini_text/gemini_image)
- Replace with per-model rate limiting using resolveAntigravityModelKey
- Remove model load statistics (IncrModelCallCount/GetModelLoadBatch)
- Simplify account selection to unified priority→load→LRU algorithm
- Remove SetAntigravityQuotaScopeLimit from AccountRepository
- Clean up scope-related UI indicators and API fields
2026-02-09 08:19:01 +08:00
erio
1af06aed96 feat: shuffle accounts within same sort group to prevent thundering herd
Add post-sort shuffle for accounts with identical (priority, loadRate,
lastUsedAt) to break deterministic ordering when concurrent requests
read the same scheduler snapshot. Applies to both Antigravity and
OpenAI scheduling paths, plus the sortAccountsByPriorityAndLastUsed
helper.

Keeps upstream CallCount/ModelLoadInfo scheduling intact; shuffle is
additive and only randomises within equivalent-rank groups.
2026-02-09 07:33:17 +08:00
erio
9236936a55 feat: route AccountTypeUpstream to ForwardUpstream in Forward() entry
Without this routing guard, ForwardUpstream is never called because
Forward() always proceeds with the standard OAuth/cookie flow.
2026-02-09 07:27:10 +08:00
erio
125152460f fix: use upstream retryDelay for rate limit duration instead of fixed default
- In handleSmartRetry, use the actual upstream retryDelay to set model
  rate limit duration instead of always using the 30s default
- Return info.RetryDelay from shouldTriggerAntigravitySmartRetry when
  shouldRateLimitModel=true, so callers know the actual delay
- Extract getDefaultRateLimitDuration() and resolveResetTime() helpers
  to reduce duplication in handleUpstreamError 429 handling
- Improve debug logging with upstream_retry_delay and response body
2026-02-09 07:11:29 +08:00
erio
6d90fb0bc3 feat: detect client disconnect during streaming and continue draining upstream for billing 2026-02-09 07:06:26 +08:00
erio
b889d5017b refactor: replace Trie-based digest session store with flat cache 2026-02-09 07:02:12 +08:00
erio
72b08f9cc5 fix: ensure sticky session failover triggers cache billing exemption 2026-02-09 06:57:07 +08:00
erio
681950dadd feat: add linear delay between Antigravity account failover switches 2026-02-09 06:56:29 +08:00
erio
a67d9337b8 feat: integrate CheckErrorPolicy into Gemini error handling paths 2026-02-09 06:55:45 +08:00
erio
2f1182e8a9 feat: unified error policy for Antigravity + enable custom error codes for Gemini accounts 2026-02-09 06:54:42 +08:00
erio
cbb4d854ab fix: check type assertion in test to satisfy errcheck linter 2026-02-09 06:47:50 +08:00
erio
35598d5648 fix: parse Gemini native request format in ParseGatewayRequest for correct session hash generation
ParseGatewayRequest only parsed Anthropic format (system/messages),
ignoring Gemini native format (systemInstruction/contents). This caused
GenerateSessionHash to produce identical hashes for all Gemini sessions.

Add protocol parameter to ParseGatewayRequest to branch between
Anthropic and Gemini parsing. Update GenerateSessionHash message
traversal to extract text from both formats.
2026-02-09 06:47:22 +08:00
erio
5c76b9e45a fix: prevent sessionHash collision for different users with same messages
Mix SessionContext (ClientIP, UserAgent, APIKeyID) into
GenerateSessionHash 3rd-level fallback to differentiate requests
from different users sending identical content.

Also switch hashContent from SHA256-truncated to XXHash64 for
better performance, and optimize Trie Lua script to match from
longest prefix first.
2026-02-09 06:46:32 +08:00
erio
0b8fea4cb4 fix: clean thoughtSignature for all clients, not just CLI
Previously, thoughtSignature cleanup only applied to Gemini CLI
requests (detected via x-gemini-api-privileged-user-id header or
tmp dir pattern). This caused 400 errors for non-CLI clients when
session cache expired and they sent stale signatures.

Remove the isGeminiCLIRequest guard so all clients benefit from
proactive thoughtSignature cleanup on session binding miss.
2026-02-09 06:45:01 +08:00
Wesley Liddick
5fa93ebdc7 Merge pull request #519 from bayma888/feature/group-sort-order
feat(admin): 新增-分组管理自由拖拽排序功能
2026-02-08 18:00:22 +08:00
bayma888
bac9e2bfd5 feat(admin): add drag-and-drop group sort order
- Add `sort_order` field to groups table with migration
- Add `PUT /api/v1/admin/groups/sort-order` API for batch update
- Implement drag-and-drop UI using vue-draggable-plus
- All queries now order groups by sort_order
- Add i18n support (en/zh) for sort-related UI text
- Update test stubs to satisfy new interface methods
2026-02-08 16:53:45 +08:00
shaw
e4d74ae11d feat(ui): 用户列表页显示当前并发数
优化 /admin/users 页面的并发数列,显示「当前/最大」格式,
参考 AccountCapacityCell 的设计风格。

- 后端 UserHandler 注入 ConcurrencyService,批量查询用户当前并发数
- 新增 UserConcurrencyCell 组件,支持颜色状态(空闲灰/使用中黄/满载红)
- 前端 AdminUser 类型添加 current_concurrency 字段
2026-02-08 16:44:51 +08:00
Wesley Liddick
2185a3b674 Merge pull request #517 from touwaeriol/fix/upstream-baseurl
refactor(upstream): replace upstream account type with apikey + auto-append base_url
2026-02-08 14:03:12 +08:00
Wesley Liddick
9e3c306a5b Merge pull request #513 from touwaeriol/pr/antigravity-full-v2
feat(antigravity): comprehensive enhancements — rate limiting, scheduling & smart retry
2026-02-08 14:01:17 +08:00
erio
69816f8691 fix: remove unused upstreamHopByHopHeaders variable to pass golangci-lint 2026-02-08 13:30:39 +08:00
shaw
b4ec65785d fix: apikey类型账号test去掉oauth-2025-04-20 2026-02-08 13:26:28 +08:00
erio
3c93644146 chore: bump version to 0.1.74.7 2026-02-08 13:14:58 +08:00
erio
fb58560d15 refactor(upstream): replace upstream account type with apikey, auto-append /antigravity
Upstream accounts now use the standard APIKey type instead of a dedicated
upstream type. GetBaseURL() and new GetGeminiBaseURL() automatically append
/antigravity for Antigravity platform APIKey accounts, eliminating the need
for separate upstream forwarding methods.

- Remove ForwardUpstream, ForwardUpstreamGemini, testUpstreamConnection
- Remove upstream branch guards in Forward/ForwardGemini/TestConnection
- Add migration 052 to convert existing upstream accounts to apikey
- Update frontend CreateAccountModal to create apikey type
- Add unit tests for GetBaseURL and GetGeminiBaseURL
2026-02-08 13:06:25 +08:00
erio
6ab77f5eb5 fix(upstream): passthrough response body directly instead of parsing SSE
ForwardUpstream/ForwardUpstreamGemini should pipe the upstream response
directly to the client (headers + body), not parse it as SSE stream.
2026-02-08 08:49:43 +08:00
erio
4f57d7f761 fix: add nil guard for gin.Context in header passthrough to satisfy staticcheck SA5011 2026-02-08 08:36:35 +08:00
erio
1563bd3dda feat(upstream): passthrough all client headers instead of manual header setting
Replace manual header setting (Content-Type, anthropic-version, anthropic-beta)
with full client header passthrough in ForwardUpstream/ForwardUpstreamGemini.
Only authentication headers (Authorization, x-api-key) are overridden with
upstream account credentials. Hop-by-hop headers are excluded.

Add unit tests covering header passthrough, auth override, and hop-by-hop filtering.
2026-02-08 08:33:09 +08:00
erio
77b66653ed fix(gateway): restore upstream account forwarding with dedicated methods
v0.1.74 merged upstream accounts into the OAuth path, causing requests
to hit the wrong protocol and endpoint. Add three upstream-specific
methods (testUpstreamConnection, ForwardUpstream, ForwardUpstreamGemini)
that use base_url + apiKey auth and passthrough the original body, while
reusing the existing response handling and error/retry logic.
2026-02-08 01:21:02 +08:00
erio
3077fd279d feat: smart retry max 1 attempt + clear sticky session on failure
- Change antigravitySmartRetryMaxAttempts from 3 to 1 to prevent
  repeated rate limiting and long waits
- Clear sticky session binding (DeleteSessionAccountID) after smart
  retry exhaustion, so subsequent requests don't hit the same
  rate-limited account
- Add flow diagrams to Forward/ForwardGemini doc comments
- Add comprehensive unit tests covering:
  - Sticky session cleared on retry failure (429, 503, network error)
  - Sticky session NOT cleared on retry success
  - Sticky session NOT cleared for non-sticky requests (empty hash)
  - Sticky session NOT cleared on long delay path (handled by handler)
  - Nil cache safety (no panic)
  - MaxAttempts constant verification
  - End-to-end retryLoop → switchError propagation with session clear
2026-02-07 19:30:58 +08:00