Kaa: Evaluating Elasticity of Cloud-Hosted DBMS

Abstract

Auto-scaling is able to change the scale of an application at runtime. Understanding the application characteristics, scaling impact as well as the workload, an auto-scaler aligns the acquired resources to match the current workload. For distributed Database Management Systems (DBMS) forming the backend of many large-scale cloud applications, it is currently an open question to what extent they support scaling at run-time. In particular, elasticity properties of existing distributed DBMS are widely unknown and difficult to evaluate and compare. This paper presents a comprehensive methodology for the evaluation of the elasticity of distributed DBMS. On the basis of this methodology, we introduce a framework that automates the full evaluation process. We validate the framework by defining significant elasticity scenarios for a case study that comprises two DBMS for write-heavy and read-heavy workloads of different intensities. The results show that scalable distributed DBMS are not necessarily elastic and that adding more instances to a cluster at run-time may even decrease the experienced performance.

A Survey on Resilience in the IoT

“…Scalability describes the capability of a software system to address a growing workload by using more resources. Accordingly, horizontal scalability denotes the capability of a distributed application to increase its performance by increasing the number of resource units (such as physical servers or virtual machines) available to the application [121]. Applications that support Elastic Scaling support the action of scaling (out or in) while the application is running, hence without downtime and ideally without impacting application performance [121].…”

“…Accordingly, horizontal scalability denotes the capability of a distributed application to increase its performance by increasing the number of resource units (such as physical servers or virtual machines) available to the application [121]. Applications that support Elastic Scaling support the action of scaling (out or in) while the application is running, hence without downtime and ideally without impacting application performance [121]. Finally, auto-scaling provides means to automatically adapt the scale-out factor of an elastically scalable application depending on current or even predicted workload.…”

An automated implementation of hybrid cloud for performance evaluation of distributed databases

“…Several studies explored these databases from an energy consumption perspective as the queries are optimized [31], or different cloud applications are leveraged [32]. Seybold et al [16] evaluated the elasticity of Cassandra and CouchDb for read and write intensive workloads in private clouds. They also presented a comprehensive methodology for evaluating the availability of Cassandra and CouchDB in case of cloud resource failure [17].…”

“…MySQL supports removing and adding node groups rather than nodes through ndb mgm create nodegroup and ndb mgm drop nodegroup commands respectively 16 . For adding a nodegroup, it is required to restart the management server and conduct a rolling restart of the existing data nodes.…”