3.8 Proceedings Paper

Extending ETSI MEC Towards Stateful Application Relocation Based on Container Migration

Publisher

IEEE
DOI: 10.1109/WoWMoM54355.2022.00035

Keywords

Edge computing; ETSI MEC; mobility; migration; QUIC

Funding

  1. Italian Ministry of Education and Research (MIUR)

Ask authors/readers for more resources

Edge computing enables running microservices near end user devices for low latency and high bandwidth applications. Mobility of devices poses a challenge to edge computing benefits, thus the need for state migration solutions. This research proposes leveraging container migration techniques to support stateful application relocation in edge computing.
Edge computing allows to run microservices in close proximity to end user devices. This proximity lets edge computing support emerging 5G application scenarios that need low latency and high bandwidth (e.g., augmented reality, autonomous vehicles). Given its interest, edge computing is fastly gaining momentum and is currently being standardised by the European Telecommunications Standards Institute (ETSI) as Multi-Access Edge Computing (MEC). Notwithstanding its strengths, edge computing is significantly challenged by device mobility, as this can reduce proximity to the edge microservice, putting edge computing benefits at risk. A way to solve this problem is to migrate the edge microservice across edge servers, to let it follow the application component running on the mobile device. Besides, if the microservice is stateful (i.e., it maintains a state associated to the user), its state needs to be migrated as well. Within ETSI MEC, this concept is expressed as stateful application relocation. The standard identifies three different high-level ways to transfer the application state. However, all of them assume that it is up to the application to actually relocate the state. In this work, we assume that applications at the edge run as containers, and we extend ETSI MEC to let it support stateful application relocation by leveraging container migration techniques. This approach allows to transfer the application state in a transparent way to the application itself. We implemented our solution and tested it over a small-scale edge computing testbed to extract initial results.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

3.8
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available