Accelerating Distributed Repartition Joins on Skewed Datasets via Patch-Based Shuffling
In distributed workloads involving joins and aggregations, skewed attribute values often cause load balancing issues, leading to stragglers and increased execution times. Existing solutions often rely on cost-based models, require extensive parameter tuning, or necessitate modifications to distribut...
Shranjeno v:
| Principais autores: | , , |
|---|---|
| Format: | Artigo |
| Jezik: | Inglês |
| Izdano: |
IEEE
2025-01-01
|
| Serija: | IEEE Access |
| Teme: | |
| Online dostop: | https://ieeexplore.ieee.org/document/10909129/ |
| Oznake: |
Brez oznak, prvi označite!
|