State Management for Cloud-Native Applications

Abstract

The stateless cloud-native design improves the elasticity and reliability of applications running in the cloud. The design decouples the life-cycle of application states from that of application instances; states are written to and read from cloud databases, and deployed close to the application code to ensure low latency bounds on state access. However, the scalability of applications brings the well-known limitations of distributed databases, in which the states are stored. In this paper, we propose a full-fledged state layer that supports the stateless cloud application design. In order to minimize the inter-host communication due to state externalization, we propose, on the one hand, a system design jointly with a data placement algorithm that places functions’ states across the hosts of a data center. On the other hand, we design a dynamic replication module that decides the proper number of copies for each state to ensure a sweet spot in short state-access time and low network traffic. We evaluate the proposed methods across realistic scenarios. We show that our solution yields state-access delays close to the optimal, and ensures fast replica placement decisions in large-scale settings.

A multi-dimensional extensible cloud-native service stack for enterprises

“…Halabi et al [48] designed a quantitative evaluation method using relative matrices to study the security properties of cloud service providers. Szalay et al [49] proposed a quantitative model to formalize the state placement problem of cloud-native applications, which can guide the architecture optimization for cloud databases. Salmon et al [50] applied classic models like End-User Computing Satisfaction (EUCS) to characterize the multifaceted properties of cloud services.…”

Ultra-Reliable and Low-Latency Computing in the Edge with Kubernetes

“…As for the technical migration overhead, we argue that stateless [26] application components can be migrated with minimal extra resources. The stateless design, of course, must be supported by a distributed cloud database [24,25], which transforms the punctual migration overhead into a continuous synchronization of application states onto multiple database instances running on nodes potentially hosting the stateless application, which leads to an extra consumption in terms of compute, memory and network resources.…”

“…However, the shared nature of certain states, and the inevitable dynamics of the application workload necessarily lead to inter-host data access within the data center (or even across data centers, if the application requires a multi-data center setup). In order to minimize the inter-host communication due to state externalization, the authors of [25,26] propose an advanced cloud scheduling algorithm that places applications' states across the hosts of a data center. In such a cloud-native setting, stateless cloud applications and an adaptively self-synchronizing distributed cloud database alleviate the long-standing issues of live migration within the cloud.…”

Design and Implementation of a Cloud PACS Architecture

“…Different message delivery models of the available brokers (e.g., at-least-once, at-most-once) provide desired properties for the system and enable robustness in case of failure. Independence from the specific cloud vendor (cloud vendor agnosticism) in the form of additional abstraction (e.g., federated storage, stateless operation) is also an important factor in today’s cloud-native design [ 53 ], permitting migration between providers without loss of functionality.…”