At MongoDB we implement the same protocols in many programming languages. How do we test that all the implementations match the specs?
Database research papers keep adding benchmarks every year, crowding out fundamental new research: stats from analyzing 800 published papers.
A half-hour talk about our enhancement to the Raft consensus protocol that strengthens Raft's consistency guarantees and time-to-recovery.
Find all the papers you added multiple times.
Closed loop benchmarks are too polite, they don't push systems past their limits.
A new Raft enhancement for fast, consistent reads.
The worst part of writing a research paper, made less bad.
I think the answer is yes, and I calculate the worst-case drift.
At MongoDB, we experimented to see if we could predict each DBaaS customer's demand fluctuations, and auto-scale them using this foreknowledge.
A multicore replication algorithm.