4534444ad8
The ordering is the load-bearing part. All three changes interlock and every wrong interleaving fails silently: flipping the usage fallback to (0, 0) before the settlement points read the derived value refunds the whole pre-deduction on success, and flipping the embedding transport before _merge goes three-state mislabels unavailable batches as measured. So the plan adds the capability first, moves all five call sites onto it while it is still equivalent, only then lets the third state take effect, and unbinds the constraint last. Registers both wiki entries and links the plan to its design.
4.6 KiB
4.6 KiB
Research Wiki 索引
自动生成,更新时间:2026-07-30 09:39 UTC
design (16)
- 2026-07-20-m1-core-design
design:2026-07-20-m1-core-design - 2026-07-20-m2-distributed-design
design:2026-07-20-m2-distributed-design - 2026-07-21-m25-resilience-design
design:2026-07-21-m25-resilience-design - 2026-07-21-m3-ocr-design
design:2026-07-21-m3-ocr-design - 2026-07-22-m4-migration-design
design:2026-07-22-m4-migration-design - 2026-07-29-settings-invariant-guards-design
design:2026-07-29-settings-invariant-guards-design - 2026-07-30-est-tokens-decoupling-design
design:2026-07-30-est-tokens-decoupling-design - 2026-07-30-settings-invariants-round-2-design
design:2026-07-30-settings-invariants-round-2-design - est_tokens 解耦: 拆分限流预扣与遥测用量兜底(issue #2)
design:est-tokens-decoupling - GatewaySettings 装配校验补齐(第二轮)
design:settings-invariants-round-2 - GatewaySettings 跨字段不变量守卫的生效范围
design:settings-invariant-guards - M1 核心里程碑设计:公共签名冻结与治理栈落地
design:m1-core-design - M2 分布式:Redis 治理后端+背压+Postgres 遥测+pricing+Embedding+压测 harness
design:m2-distributed - M2.5 治理韧性: 半死源隔离与健康感知调度
design:m25-resilience - M3 OCR 端口族设计
design:m3-ocr - M4 迁移验证设计(GovDoc→CHS,发 v1.0)
design:m4-migration
finding (11)
- 2026-07-20-m2-soak-workload
finding:2026-07-20-m2-soak-workload - 2026-07-21-m25-acceptance
finding:2026-07-21-m25-acceptance - 2026-07-21-p6-soak-baseline
finding:2026-07-21-p6-soak-baseline - 2026-07-22-m4-acceptance
finding:2026-07-22-m4-acceptance - 2026-07-22-p7-ocr-soak
finding:2026-07-22-p7-ocr-soak - M2 verifier 三项 Important 补齐(不变量接线/网关保护/P3 验收)
finding:m2-verifier-fixes - M2 真实数据压测: 场景矩阵与数据清单
finding:m2-soak-workload - M2.5 验收: P6 同场景 58.1% → 98.96%
finding:m25-acceptance - M4 迁移验收(GovDoc+CHS)
finding:m4-acceptance - P6 混合浸泡首跑基线与记分板三重伪击穿修复
finding:p6-soak-baseline - P7 OCR soak 验收: 99.73% 与 13 不变量全 PASS
finding:p7-ocr-soak
plan (12)
- 2026-07-20-m1-core-plan
plan:2026-07-20-m1-core-plan - 2026-07-20-m2-distributed-plan
plan:2026-07-20-m2-distributed-plan - 2026-07-21-m25-resilience-plan
plan:2026-07-21-m25-resilience-plan - 2026-07-21-m3-ocr-plan
plan:2026-07-21-m3-ocr-plan - 2026-07-22-m4-migration-plan
plan:2026-07-22-m4-migration-plan - 2026-07-30-est-tokens-decoupling-plan
plan:2026-07-30-est-tokens-decoupling-plan - est_tokens 解耦实施计划
plan:est-tokens-decoupling - M1 核心里程碑实现计划
plan:m1-core-plan - M2 分布式实现计划
plan:m2-distributed - M2.5 治理韧性实现计划
plan:m25-resilience - M3 OCR 实现计划
plan:m3-ocr - M4 迁移实现计划(T0-T14)
plan:m4-migration
schema (1)
- 表结构: llm_calls(遥测 18 字段)
schema:llm-calls
metric (2)
- OCR 治理调用成功率与错误分类分布
metric:ocr-call-success - 每次调用必录覆盖率(含缓存命中/失败/取消)
metric:call-telemetry-coverage