Apache Spark has been widely adopted for in-memory data analytics at scale, however, efficient memory utilization is a common challenge, and users will either run out of memory or experience low and unstable performance. Many Spark users may not be aware of the differences in memory utilization between caching data directly in-memory into the Spark JVM versus storing data off-heap via an in-memory storage service like Alluxio. In this office hour, I will highlight the two approaches with a demo and open up for discussions
In this Office Hour we’ll go over:
- How to run Spark shell with Alluxio such that Spark jobs
- A demo to compare the memory usage between Spark cache and using Alluxio as the external off-heap caching service
- Open Session for discussion on any topics such as running Presto on Alluxio, and more
ALLUXIO COMMUNITY OFFICE HOUR
Apache Spark has been widely adopted for in-memory data analytics at scale, however, efficient memory utilization is a common challenge, and users will either run out of memory or experience low and unstable performance. Many Spark users may not be aware of the differences in memory utilization between caching data directly in-memory into the Spark JVM versus storing data off-heap via an in-memory storage service like Alluxio. In this office hour, I will highlight the two approaches with a demo and open up for discussions
In this Office Hour we’ll go over:
- How to run Spark shell with Alluxio such that Spark jobs
- A demo to compare the memory usage between Spark cache and using Alluxio as the external off-heap caching service
- Open Session for discussion on any topics such as running Presto on Alluxio, and more
Video:
Presentation slides:
ALLUXIO COMMUNITY OFFICE HOUR
Apache Spark has been widely adopted for in-memory data analytics at scale, however, efficient memory utilization is a common challenge, and users will either run out of memory or experience low and unstable performance. Many Spark users may not be aware of the differences in memory utilization between caching data directly in-memory into the Spark JVM versus storing data off-heap via an in-memory storage service like Alluxio. In this office hour, I will highlight the two approaches with a demo and open up for discussions
In this Office Hour we’ll go over:
- How to run Spark shell with Alluxio such that Spark jobs
- A demo to compare the memory usage between Spark cache and using Alluxio as the external off-heap caching service
- Open Session for discussion on any topics such as running Presto on Alluxio, and more
Video:
Presentation slides:
Videos:
Presentation Slides:
Complete the form below to access the full overview:
.png)
Videos
In the rapidly evolving landscape of AI and machine learning, Platform and Data Infrastructure Teams face critical challenges in building and managing large-scale AI platforms. Performance bottlenecks, scalability of the platform, and scarcity of GPUs pose significant challenges in supporting large-scale model training and serving.
In this talk, we introduce how Alluxio helps Platform and Data Infrastructure teams deliver faster, more scalable platforms to ML Engineering teams developing and training AI models. Alluxio’s highly-distributed cache accelerates AI workloads by eliminating data loading bottlenecks and maximizing GPU utilization. Customers report up to 4x faster training performance with high-speed access to petabytes of data spread across billions of files regardless of persistent storage type or proximity to GPU clusters. Alluxio’s architecture lowers data infrastructure costs, increases GPU utilization, and enables workload portability for navigating GPU scarcity challenges.
In this talk, Zhe Zhang (NVIDIA, ex-Anyscale) introduced Ray and its applications in the LLM and multi-modal AI era. He shared his perspective on ML infrastructure, noting that it presents more unstructured challenges, and recommended using Ray and Alluxio as solutions for increasingly data-intensive multi-modal AI workloads.
As large-scale machine learning becomes increasingly GPU-centric, modern high-performance hardware like NVMe storage and RDMA networks (InfiniBand or specialized NICs) are becoming more widespread. To fully leverage these resources, it’s crucial to build a balanced architecture that avoids GPU underutilization. In this talk, we will explore various strategies to address this challenge by effectively utilizing these advanced hardware components. Specifically, we will present experimental results from building a Kubernetes-native distributed caching layer, utilizing NVMe storage and high-speed RDMA networks to optimize data access for PyTorch training.