Ilyas Toumlilt

It is my pleasure to invite you to the livestream of my thesis defense, completed within the LIP6 Delys INRIA team and titled "Colony: A Hybrid Consistency System for Highly-Available Collaborative Edge Computing".
The defense will be in english and take place on Tuesday, Decembre 21, 2021 at 13:00 (Paris Time).
The livestream video will be accessible, minutes before the defense, from this page.

Edge applications, such as gaming, cooperative engineering, or in-the-field information sharing, enjoy immediate response, autonomy and availability by distributing and replicating data at the edge. However, application developers and users demand the highest possible consistency guarantees, and specific support for group collaboration. To address this challenge, Colony guarantees Transactional Causal Plus Consistency (TCC+) globally, strengthened to Snapshot Isolation within edge groups. To help with scalability, fault tolerance and security, its logical communication topology is forest-like, with replicated roots in the core cloud, but with the flexibility to migrate a node or a group. Despite this hybrid approach, applications enjoy the same semantics everywhere in the topology. Our experiments show that local caching and peer groups improve throughput and response time significantly, performance is not affected in offline mode, and that migration is seamless.

Keywords : Causal Consistency, Collaborative Computing, Edge Computing, CRDTs, Peer-to-Peer Systems.

Much previous work on data in edge computing focuses on streaming and content delivery.
Examples include sensor systems or propagating database views. We leverage this previous work by propagating shared state as a stream of update events. The distributed sharing of persistent mutable state raises extra challenges, which we address in this state of the art study.
Let’s study in detail each of these systems.

Keywords : Geo-replication, Partial Replication, Edge Computing, Causal Consistency, Peer-to-peer.

Cloud-scale services improve availability and latency by geo-replicating data in several data centers (DC) across the world. Nevertheless, the closest DC is often still too far away for an optimal user experience. To remain available at all times, client-side applications need to cache data at client machines, caching data at client machines can improve availability and latency for many applications, and also allow for temporary disconnection. This approach is used in many recent cloud services, like Google Drive RT API or Mobius [3, 5, 9, 16], where developers implement caching and buffering at application level, but it doesn’t ensure system-wide consistency guarantees.

Keywords : Geo-replication, Partial Replication, Edge Computing, Causal Consistency, Peer-to-peer.

Pour augmenter les performances et la disponibilité, les bases de données sont généralement répliquées sur différents serveurs au sein de centres de données à travers le monde. Cependant, la réplication implique quelques problèmes de cohérence. La cohérence causale est un choix intéressant pour construire ce modèle de bases de données. Elle permet aux clients d'observer un état cohérent par rapport à leurs écritures, en incluant les opérations qu'ils ont précédemment observé, tout en minimisant les anomalies pour le programmeur. Cependant, les mises en œuvre actuelles de la cohérence causale sont susceptibles de connaitre des ralentissements en chaîne à l'échelle de plusieurs centres de données. Ce papier décrit la conception d'un système causalement cohérent, évolutif, et non exposé à l'impact que peut avoir un serveur lent ou défaillant, sur la disponibilité des données.

Mots-clés : Bases de données distribuées, géo-replication, cohérence causale, disponibilité sous partition.