SWARM: Adaptive Load Balancing in Distributed Streaming Systems for Big Spatial Data

[Sunt-ing]

https://raw.githubusercontent.com/Sunt-ing/database-system-readings/main/papers/SWARM.pdf

[Sunt-ing]

What problem does the paper solve? Is it important?

SWRAM is a lightweight dynamic load-balance protocol for distributed spatial data streaming systems. Importance: Probably yes. Nowadays people tend to use simple load balance methods, but this protocol can be a good try.

How does it solve the problem?

Cost model: C(p) = N(p) ×Q(p) ×Prob(p), where N means the number of data points, Q means the number of queries.

What could be improved?

• I'd like to answer what can be gained: C(p) = N(p) ×Q(p) ×Prob(p) is often used. But for simplicity, C(p) = N(p) ×Q(p) can be used first.

Others

• Published by Walid G. Aref as it was, this paper is of low quality. Almost everything in this paper is well known in the literature.
• Papers in journals are usually too tedious and of little novelty. Take SWARM as an example, the paper is as long as 43 pages!
• The pictures in this paper are good: beautiful and illustrative.