95% Faster Cold Start times - For any AI/ML/Large Container

Managing software distribution at scale is a significant challenge for enterprises. Specifically, applications cannot commence until they are deployed and installed on a node. Additionally, when deploying large numbers of applications to thousands of machines locally, globally, or into HPC compute clusters, network load issues arise. For example, deploying a 1GB application to 1000 nodes requires transmitting 1TB over your network! Considering that AI/ML applications typically are larger than 4GB and can easily exceed 10GB this can be a considerable improvement.

While container images can be seen as another method of software distribution, containerizing applications can exacerbate this issue. Therefore, container-based software distribution at scale issues must be comprehensively understood and addressed.

In this one-hour free webinar, you will learn about Alternate Image Storage and Alternate Layer Storage capabilities available today in Docker/containers and Podman/CRI-O. You will understand how these technologies function and how they help alleviate traditional software distribution at scale issues, including emerging concerns such as the need for fast spin-up of microservices.

Gerry will also outline how a major financial institution leverages these technologies in conjunction with an AFS Family Distributed file system to distribute applications globally to tens of thousands of nodes.

About the Instructor:

Gerry Seidman has designed and implemented numerous high-performance distributed systems. He currently serves as the President at AuriStor and leads the design and implementation of all AuriStor container/Kubernetes/OpenShift products.