Join Algorithms Meet New Constraints: Scalability, Latency and Privacy
summary
Algorithms that scale joins across hundreds or thousands of machines by minimizing communication rounds and per-server load, with formal claims of an optimal rounds-vs-load trade-off; dynamic data structures that trade preprocessing and per-update cost against worst-case enumeration delay so joins can be updated and streamed with provable latency bounds; privacy-preserving execution that hides communication/read-write patterns via differentially oblivious algorithms and methods for private synthetic join results.
AI-written summary · about
related talks
39mMassively Parallel Evaluation of Relational Queries II
01h 20mAn Introduction to Distributed, Massively Parallel, and Local Computation Algorithms
31mSome Novel Computational Models to Think About
26mLarge-Scale Generalized Matching
27mMetadata-private data transfer
19mLook Before, Before You Leap: Online Vector Load Balancing with Few Reassignments
38mHow Scale Makes Distributed Systems Slower
25mAvoiding Overload in Distributed Systems