Comparison of Autoscaling Frameworks for Containerised Machine-Learning-Applications in a Local and Cloud Environment

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Schroeder, Christian, Boehm, Rene, Lampe, Alexander
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929255665893376
author Schroeder, Christian
Boehm, Rene
Lampe, Alexander
author_facet Schroeder, Christian
Boehm, Rene
Lampe, Alexander
contents When deploying machine learning (ML) applications, the automated allocation of computing resources-commonly referred to as autoscaling-is crucial for maintaining a consistent inference time under fluctuating workloads. The objective is to maximize the Quality of Service metrics, emphasizing performance and availability, while minimizing resource costs. In this paper, we compare scalable deployment techniques across three levels of scaling: at the application level (TorchServe, RayServe) and the container level (K3s) in a local environment (production server), as well as at the container and machine levels in a cloud environment (Amazon Web Services Elastic Container Service and Elastic Kubernetes Service). The comparison is conducted through the study of mean and standard deviation of inference time in a multi-client scenario, along with upscaling response times. Based on this analysis, we propose a deployment strategy for both local and cloud-based environments.
format Preprint
id arxiv_https___arxiv_org_abs_2311_18659
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Comparison of Autoscaling Frameworks for Containerised Machine-Learning-Applications in a Local and Cloud Environment
Schroeder, Christian
Boehm, Rene
Lampe, Alexander
Distributed, Parallel, and Cluster Computing
94-04
I.2.11
When deploying machine learning (ML) applications, the automated allocation of computing resources-commonly referred to as autoscaling-is crucial for maintaining a consistent inference time under fluctuating workloads. The objective is to maximize the Quality of Service metrics, emphasizing performance and availability, while minimizing resource costs. In this paper, we compare scalable deployment techniques across three levels of scaling: at the application level (TorchServe, RayServe) and the container level (K3s) in a local environment (production server), as well as at the container and machine levels in a cloud environment (Amazon Web Services Elastic Container Service and Elastic Kubernetes Service). The comparison is conducted through the study of mean and standard deviation of inference time in a multi-client scenario, along with upscaling response times. Based on this analysis, we propose a deployment strategy for both local and cloud-based environments.
title Comparison of Autoscaling Frameworks for Containerised Machine-Learning-Applications in a Local and Cloud Environment
topic Distributed, Parallel, and Cluster Computing
94-04
I.2.11
url https://arxiv.org/abs/2311.18659