Apr 6, 2025 · Databases

Accelerating StarRocks Node Decommissioning via tablet_sched_slot_num_per_path Tuning

The author decommissioned three StarRocks nodes with over 1M tablets each; default replica migration moved only 200k replicas in 24 hours, so they increased tablet_sched_slot_num_per_path from 2 to 32 per disk during off-peak, achieving 800k+ replicas migrated per node in 8 hours while monitoring disk I/O.

Database AdministrationStarRocksnode decommissioning

0 likes · 4 min read

Accelerating StarRocks Node Decommissioning via tablet_sched_slot_num_per_path Tuning

Architects' Tech Alliance

Apr 1, 2023 · Industry Insights

Why GPUs Lag Behind Big AI Models and How In‑Memory Computing Helps

The article examines the growing bottlenecks of large‑scale AI model training caused by the separation of storage and compute, analyzes why conventional GPU architectures cannot keep pace with exponential model growth, and presents in‑memory and near‑memory computing, as well as storage‑compute integration, as promising solutions to boost performance, energy efficiency, and scalability for cloud and edge deployments.

AI ComputeCloud ComputingEnergy Efficiency

0 likes · 10 min read

Why GPUs Lag Behind Big AI Models and How In‑Memory Computing Helps

storage-compute integration

Accelerating StarRocks Node Decommissioning via tablet_sched_slot_num_per_path Tuning

Why GPUs Lag Behind Big AI Models and How In‑Memory Computing Helps