今日固定时间表
三块窗口固定：08:50-10:55 系统设计，14:30-16:15 算法，20:30-21:15 口述。
| 08:50-09:15 | 08:50-10:55 system design Bad/Good/Great - Bad: 长事务锁座；Good: status+expiry；Great: durable hold+version+TTL+idempotent checkout。 |
|---|---|
| 09:15-09:50 | 08:50-10:55 system design 精确原文 - 读 Ticketmaster Potential Deep Dives、5 个 deep-dive anchors 和 level expectations。 |
| 09:50-10:25 | 08:50-10:55 system design 故障时序 - TTL 到期同时 payment success、重复 webhook、两个用户抢同 seat、waiting room token replay。 |
| 10:25-10:55 | 08:50-10:55 system design 恢复和指标 - oversell invariant、hold expiry lag、duplicate rejection、checkout reconciliation gap、event admission rate。 |
| 14:30-16:15 | NeetCode Tag · Binary Search 三道题，每题 30m solve + 5m evidence，不能改 tag。 |
| 20:30-21:15 | 20:30-21:15 Staff Q&A：reservation expiry, lock alternatives, stale seat map, payment webhook retry。 |
算法模块
连续 NeetCode Binary Search；每天 exactly 3 slots。
| # | Problem / mode | Pattern | Invariant | Bug risk | Time | Space |
|---|---|---|---|---|---|---|
| 1 | Arranging Coins NEW · 30m solve + 5m evidence | binary search answer | All k with k(k+1)/2 <= n are feasible. | Integer overflow in triangular sum; off-by-one on last feasible. | O(log n) | O(1) |
| 2 | Valid Perfect Square NEW · 30m solve + 5m evidence | integer square predicate | lo..hi contains every candidate root not yet disproved. | Using floating sqrt; product overflow. | O(log n) | O(1) |
| 3 | Sqrt(x) NEW · 30m solve + 5m evidence | floor binary search | best is the largest mid where mid*mid <= x. | Returning first infeasible rather than best feasible. | O(log x) | O(1) |
Hello Interview 精确目录
保留 live canonical heading title；day page 使用 direct section anchors。
| 项目 | 精确 canonical section | Disposition | 今天怎么用 |
|---|---|---|---|
| Ticketmaster | Potential Deep Dives | deep dive | move from functional HLD to contention and high demand |
| Ticketmaster | 1) How do we improve the booking experience by reserving tickets? | deep dive | seat hold, TTL, expiry, checkout and recovery |
| Ticketmaster | 2) How is the view API going to scale to support 10s of millions of concurrent requests during popular events? | deep dive | cache static event data while keeping booking authoritative |
| Ticketmaster | 3) How will the system ensure a good user experience during high-demand events with millions simultaneously booking tickets? | deep dive | seat-map staleness, SSE updates, waiting room, admission control |
| Ticketmaster | 4) How can you improve search to ensure we meet our low latency requirements? | deep dive | indexed or search-engine-backed event search |
| Ticketmaster | 5) How can you speed up frequently repeated search queries and reduce load on our search infrastructure? | deep dive | query result caching and edge caching |
| Ticketmaster | What is Expected at Each Level? | level expectation | calibrate Senior+/Staff bar |
| Ticketmaster | Mid-level | level expectation | functional breadth and basic no-double-booking |
| Ticketmaster | Senior | level expectation | deep search, reservation, high-demand trade-offs |
| Ticketmaster | Staff+ | level expectation | proactive high-contention, failure and business-policy decisions |
DDIA 精确决策卡
只使用 Ch8 和 Ch10 中直接解决今天设计决定的小节。
| 项目 | 项目问题 | 精确 DDIA subsection | 改变的系统决定 |
|---|---|---|---|
| Ticketmaster | What consistency is required so one unique seat cannot be sold twice? | Ch10 - 线性一致性 | The seat object must expose one linearization point for held->sold; search and seat-map reads can remain stale projections. |
| Ticketmaster | Which isolation anomalies would break no-oversell? | Ch8 - 防止丢失更新 | Lost update and stale read-modify-write are not acceptable on seat/hold rows; use row lock, OCC, or serializable/conditional update. |
三域 trace 任务
Winning transaction标出唯一线性化点：seat hold/sold、accepted bid at close、brokerage order/execution state。
Duplicate/retry failure给出 stable idempotency key、重复返回语义、旧 owner/stale version 被拒绝的证据。
Projection boundary说明哪些读可以陈旧：seat map、highest bid display、market price UI；哪些写不能陈旧。
本日 artifacts
项目脚本、音频、PDF 和 mock evidence 从当天直接打开。
今日产出、修复与英语 recall
| 课程角色 | CORE · all deep dives + Bad/Good/Great |
|---|---|
| 必须交付 | Ticketmaster deep-dive matrix：reservation, read scale, high-demand UX, search, query caching 各含 invariant/failure/recovery/metrics。 |
| 修复规则 | 如果答案只说 Redis lock，补 durable hold row、version、expiry、payment reconciliation 和 refund compensation。 |
| English recall | English recall: The waiting room controls admission; it is not inventory ownership. |
Detailed lecture notes, audio, recall scripts, PDFs, Staff Q&A, and mock packs are archived locally and are intentionally not published on this site.