Using CheatSheets To Apply Best Practices

CheatSheet: Feature Design For Job Interview

CheatSheet: Feature Design For Job Interview

1.1 Top 20 Design Problems For Technical Modules

Num Name Summary
1 Design a distributed counter link, link
2 Delayed task scheduling link
3 Design a thread-safe Hashmap link, link
4 Design a distributed Hashmap  
5 Design a distributed UUID generator  
6 Design An API Rate Limiter link, link, link
7 Design a distributed transaction  
8 Top URL hits  
9 Unique url hits  
10 Design a distributed transactions  
11 Design a load balancer  
12 Design a client-server API to build a rich document editor  
13 Design online/offline status system  
14 Design a circuit breaker  
15 Design a secrets management system  
16 Design data sync for a distributed system  
17 Design: A Parking Lot Service link
18 Design: A URL Redirecting Feature  
19 Design a service auto-discovery feature  
20 Design A big file transfer feature  
21 TODO  

1.2 Top 30 Concepts For Feature/System Design

Num Name Summary
1 Caching Stores data so that future requests of data retrieval can be faster
2 Message Queue Provides an asynchronous communications protocol,
3 Data Partition & Sharding Break up a big data volume into many smaller parts
4 DB Indexing Create indexes on multiple columns to speed up table look up
5 DB replication Duplicate data to increase service availability
6 CAP: Consistency/Availability/Partition A distributed database system can only have 2 of the 3
7 DB: SQL & NoSQL Relational databases and non-relational databases
8 Concurrency & Communication  
9 Pessimistic And Optimistic Locking  
10 Consistency Module weak consistency, eventual consistency, strong consistency
11 Conflict resolution Quorum, vector lock, reconcile on read/write, CRDT
12 Networking: HTTP  
13 Pull vs Push model  
14 Garbage Collection  
15 Memory Management  
16 Heartbeats  
17 Self Protection API Rate limit, Circuit breaker, bulkhead, throttling
18 Filesystem  
19 API: gRPC vs REST  
20 Load balancer  
21 Scale up vs Scale out Vertical scaling and Horizontal scaling
22 API Design  
23 Session management  
24 Networking: TCP vs UDP  
25 Consistency patterns Weak consistency, Eventual consistency, Strong consistency
26 Availability patterns Fail-over vs Replication
27 CDN – Content Delivery Network Edge caching
28 Monitoring  
29 Security  
30 Networking: DNS  

1.3 Advanced Data Structure & Algorithms

Num Name Summary
1 Consistent Hash  
2 Delayed queue Run scheduled tasks
3 Bloom filter A space-effcient query returns either “possibly in set” or “definitely not”
4 Gossip Propagate cluster status
5 CRDT(Conflict-Free Replicated Data Types)  
6 SSTable (Sorted Strings Table)  
7 LSM (Log Structured Merge Trees)  
8 Two-phase commit/Three-phase commit  
9 B+ Tree  
10 Vector Clocks/Version Vectors  
11 Paxos and raft protocol  
12 Merkle Tree  

CheatSheet: Feature Design For Job Interview

1.4 Explain workflow: What happens when XXX?

Num Name Summary
1 When happens when I search in google?  
2 How loadbalancer works  
3 Explain three phase commit model  
4 Explain HTTP return code  
5 Explain Mysql DB replication model  
6 Explain gossip protocol  
7 Explain CAP  
8 Explain Hadoop file system  

1.5 Explain tools: how XXX supports XXX?

Num Name Summary
1 How JDK implement hashmap?  
2 Explain java garbage collection model  
3 Explain raft/etcd  
4 How OS supports XXX?  

1.6 Cloud Design Principles

Num Name Summary
1 Fail fast  
2 Design for failure  
3 Immutable infrastructure  
4 Cats vs Cattle Avoid snowflake servers
5 Auto healing  
6 Async programming  
7 GitOps operational model  
8 Event-Driven Architectures  

1.7 Cloud Design Patterns

Num Name Summary
1 Ambassador pattern Create helper service to send network requests, besides the main sevice
2 Cache-Aside pattern Load data on demand into a cache from a data store
3 Circuit Breaker pattern If a request takes too many reousrce, abort it
4 Bulkhead pattern Isolate elements into pools, so that one fire won’t burn all
5 Gateway Aggregation pattern Aggregate multiple individual requests into a single request
6 Priority Queue pattern Support different SLAs for different individual clients
7 Strangler pattern Incrementally migrate a legacy system piece by piece

1.8 Misc

Num Name Summary
1 How to store 2TB data into 3 disks of 1TB. And be tolerant for one disk failure A, B, C. And C = A XOR B
2 Find out the difference between two files. Majority of these two are the same #lcs – Longest Common Subsequence
3 How to support feature of “diff 1.txt 2. txt”  
4 Avoid double payment in a distributed payment system link


Leave a Reply

Your email address will not be published. Required fields are marked *