Commit Graph
6 Commits
Author SHA1 Message Date
wangbo 8485f14bc4 test(acceptance): 验证厂商三类额度与优先路由
为 provider-burst 固定三平台并发、RPM、TPM 和优先级组合,使用稳定 token 用量的本地视频负载验证同级按容量分流及低优先级溢出。\n\n报告升级为 v2,强制校验三类额度峰值精确命中上限、不越界、队列回收、三 Worker 分发以及重复提交和回调安全。\n\n验证:\n- env -u AI_GATEWAY_TEST_DATABASE_URL go test ./... -count=1\n- go vet ./...\n- bash -n scripts/acceptance/provider-burst.sh scripts/acceptance/run-local-acceptance.sh\n- shellcheck -x scripts/acceptance/provider-burst.sh scripts/acceptance/run-local-acceptance.sh
2026-08-03 11:48:54 +08:00
wangbo 6c5daf29ca perf(queue): 按策略动态扩缩异步 Worker
将 River 执行容量按平台模型与用户组的有效并发策略动态调整,并为长任务续租、并发租约原子抢占和限流退避补充保护。\n\n统一平台模型限流继承语义,兼容历史 platformLimits/modelLimits,并为三个迁移图像模型建立独立并发租约。补充管理端显式继承/覆盖配置、指标、单元测试及隔离 PostgreSQL 验收。\n\n验证:go test ./... -count=1;pnpm lint;pnpm test;pnpm build;隔离 PostgreSQL 并发原子性/续租测试;128 任务与三分钟长任务动态 Worker 验收。
2026-07-24 12:26:56 +08:00
wangbo 0e675e259c feat: add runner failover rules and cache affinity 2026-06-28 20:50:23 +08:00
wangbo 7f32446466 feat: support cached input token billing 2026-06-23 17:17:57 +08:00
wangbo 71950d2b4f fix(rate-limits): treat empty policies as unlimited 2026-05-24 22:28:25 +08:00
wangbo ba850a06c6 feat: improve model rate limit tracking 2026-05-12 03:22:29 +08:00