Skip to content

title: Cheatsheets description: Dense reference tables for the night before: latency numbers, selection matrices, consistency tables, the problem-to-pattern map, checklists and a glossary.


Cheatsheets

Dense, scannable tables for the night before and the minutes between rounds: the canonical latency and estimation numbers every page on this site uses, database and queue selection matrices, consistency and isolation tables, the problem-to-pattern map, round checklists, the clarifying-question bank and a glossary. They summarise; the linked pages explain. Keep the latency sheet open while you practise estimation until the orders of magnitude are automatic.

Tiers: P0 must be solid before any interview, P1 is expected of a strong SDE2, P2 is stretch material for when the rest is done.

Start here (P0)

All pages

# Page Tier Covers
1 Latency numbers and estimation tables P0 Latency ladder (ranges), powers of two, nines to downtime, QPS/storage/bandwidth formulas, per-node capacity rules of thumb, object sizes…
2 Database selection matrix P0 Rows: PostgreSQL/MySQL, DynamoDB, Cassandra, MongoDB, Redis, Elasticsearch, Neo4j, InfluxDB/Timescale, ClickHouse/BigQuery…
3 Queue and stream selection P1 Kafka, RabbitMQ, SQS/SNS, Pulsar, Redis Streams, Kinesis by ordering, delivery, retention and replay, throughput, fan-out, use when.
4 Consistency, replication and isolation tables P1 Replication topology by consistency, failover and conflicts; isolation levels by anomalies; N/W/R configurations and their guarantees…
5 Problem to pattern quick reference P0 Symptom to pattern to Pythonic idiom to example problem page; confusion pairs (Strategy/State/Template Method, Decorator/Proxy/Adapter…
6 HLD round checklist P0 Step-by-step with must-say phrases, required artifacts per step, red flags, ten deep-dive starters that work for any system.
7 LLD round checklist P0 Step-by-step, Amazon OOD rubric mapping, code-quality checklist (types, dataclasses, enums, exceptions, tests), when to stop adding…
8 Common SDE2 mistakes in design rounds P1 Twenty HLD and twenty LLD mistakes, each with symptom, why it costs, fix.
9 Clarifying-question bank P1 Grouped: users and actors, scale, read/write ratio, consistency, latency, durability, availability, geography, cost, security and privacy…
10 Glossary P1 At least 150 terms, one line each, each linked to the page that defines it.
11 Off-the-shelf building blocks P1 One-paragraph cards: Redis, Memcached, Kafka, RabbitMQ, Cassandra, DynamoDB, PostgreSQL, Elasticsearch, ZooKeeper/etcd, S3, CDN…