Increasing the Execution Speed of Containerized Analysis Workflows Using an Image Snapshotter in Combination With CVMFS

Abstract

The past years have shown a revolution in the way scientific workloads are being executed thanks to the wide adoption of software containers. These containers run largely isolated from the host system, ensuring that the development and execution environments are the same everywhere. This enables full reproducibility of the workloads and therefore also the associated scientific analyses performed. However, as the research software used becomes increasingly complex, the software images grow easily to sizes of multiple gigabytes. Downloading the full image onto every single compute node on which the containers are executed becomes unpractical. In this paper, we describe a novel way of distributing software images on the Kubernetes platform, with which the container can start before the entire image contents become available locally (so-called “lazy pulling”). Each file required for the execution is fetched individually and subsequently cached on-demand using the CernVM file system (CVMFS), enabling the execution of very large software images on potentially thousands of Kubernetes nodes with very little overhead. We present several performance benchmarks making use of typical high-energy physics analysis workloads.

Scalable Declarative HEP Analysis Workflows for Containerised Compute Clouds

“…We did not observe issues with the creation of these large container images and storing them on either the Docker Hub (DOC, 2021) or the CERN GitLab registries. Further, the pulling of containers can be greatly reduced by making use of "lazy pulling", an approach through which only those parts of the image actually required for the execution of the workload are downloaded (Mosciatti et al, 2021). Secondly, we evaluated a more general point of whether it is possible (and advantageous) to express the complex computational processes inherent in experimental particle physics data analyses in the form of declarative workflow languages, abstracting the details of the control flow.…”