Overview
What this challenge is about.
Build an LSM-tree engine in Rust with two compaction strategies, run three workloads, and earn your verifiable certificate.
The scenario
The institute is a publicly funded systems-research lab in Zurich with 14 researchers, building open teaching material so graduate students can study real storage-engine trade-offs hands-on rather than from textbook diagrams.
The Brief
What you'll do, and what you'll demonstrate.
Complete a teaching-grade Rust LSM-tree storage engine with pluggable size-tiered and leveled compaction, and determine from your own benchmarks which strategy suits which workload.
Earning criteria — what you'll demonstrate
- Implement the core components of an LSM-tree engine (write-ahead log, memtable, SSTables, bloom filters, sparse index) and reason about how they interact on the read and write paths
- Guarantee durability and recover a consistent state after a crash through write-ahead log replay under injected faults
- Implement size-tiered and leveled compaction behind a common interface and explain how each reshapes data on disk
- Design and run YCSB-shaped workloads and measure write amplification, read latency percentiles, and space amplification rigorously
- Translate measurements into a defensible engineering recommendation about compaction strategy selection
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Aligned coursework coming soon.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
- Lsm Tree
Apply lsm tree to solve real industry problems and demonstrate production-level capability.
- Storage Engine
Apply storage engine to solve real industry problems and demonstrate production-level capability.
- Rust
Apply rust to solve real industry problems and demonstrate production-level capability.
- Compaction Strategies
Apply compaction strategies to solve real industry problems and demonstrate production-level capability.
- Ycsb
Apply ycsb to solve real industry problems and demonstrate production-level capability.
- Crash Recovery
Apply crash recovery to solve real industry problems and demonstrate production-level capability.
Careers
Career paths this challenge builds toward
Completing this challenge demonstrates skills that transfer directly to these roles:
Storage Engine Engineer
Building a working LSM-tree from write-ahead log to SSTables mirrors the daily work of engineers on database and key-value store teams, where durability, recovery, and compaction trade-offs are the core of the job.
This challenge sharpens
- lsm-tree
- storage-engine
- crash-recovery
Systems Software Engineer (Rust)
Implementing a performance-critical engine in Rust with careful concurrency and on-disk formats builds the exact competencies systems teams hire for in infrastructure and platform roles.
This challenge sharpens
- rust
- storage-engine
- compaction-strategies
Database Performance Engineer
Designing YCSB-shaped workloads and turning write/space-amplification and latency measurements into strategy recommendations is the heart of performance engineering for data systems.
This challenge sharpens
- ycsb
- compaction-strategies
- lsm-tree