Nov 11, 2022 · Big Data

How Huolala Boosted Offline Scheduling Performance: Strategies & Lessons

Huolala’s big‑data offline platform, built from scratch, faced escalating scheduling delays as task instances grew, prompting a series of short‑ and mid‑term optimizations—including zombie task cleanup, retention policies, memory caching, algorithmic tweaks, and high‑availability enhancements—to dramatically reduce dependency computation time and sustain million‑scale daily workloads.

Big DataDistributed Systemsoffline scheduling

0 likes · 12 min read

How Huolala Boosted Offline Scheduling Performance: Strategies & Lessons

DataFunTalk

Jun 12, 2022 · Big Data

Huya Offline Job Scheduling System: Design, Baseline Scheduling, and Cost Optimization

This article introduces Huya's offline job scheduling platform, covering its positioning, evolution, system architecture, baseline scheduling techniques, cost‑optimization strategies, resource‑balancing methods, and future intelligent data‑warehouse directions, illustrating how data‑driven automation improves YARN utilization and SLA compliance.

Cost OptimizationDAGYARN

0 likes · 12 min read

Huya Offline Job Scheduling System: Design, Baseline Scheduling, and Cost Optimization