Automatic provisioning of multi-tier applications in cloud computing environments

Abstract

Provisioning of multi-tier applications in cloud environments raises new challenges not addressed by prior work on provisioning single-tier applications, on dynamic balancing or on resource allocation in other types of distributed systems. Flexible and general automatic mechanisms are needed to determine how much virtual resources need to be allocated to each tier of the application minimizing resources consumption and meeting the service level agreement. Both the research community and the main cloud providers are proposing this kind of solutions but most of them are application-specific, provider-specific, centralized and focused only on batch applications. This paper presents an automatic provisioning solution for multi-tier applications called AutoMAP. The proposed mechanism is general (application and provider independent), it can be implemented with different architectures from centralized to distributed even being provided as a service, and it is able to deal with both batch and interactive applications allowing horizontal and vertical scaling (based on replication and on resizing respectively). A first prototype of AutoMAP has been implemented to demonstrate its efficiency with experimental results using a widely used benchmark, RUBiS, on a real cloud architecture.

Chamulteon: Coordinated Auto-Scaling of Micro-Services

“…In 2015, Beltrán presented AutoMAP [7], a reactive autoscaler. Its provisioning model is based on response time triggers, so it implements a rule-based approach.…”

“…Examples from the first category include Chameleon, which we proposed in our previous work [2], and the famous open-source auto-scalers React, Adapt, Hist, and Reg [3]- [6]. In contrast, popular auto-scalers from the second category, such as AutoMap, AGILE, and CloudScale [7]- [9], are closed-source.…”

BECloud: A new approach to analyse elasticity enablers of cloud services

“…Only elasticity enablers on the user's side can be used (BECloud is not able to change provider's design, deployment or management decisions), as it will be discussed in the next section, mainly problem, service and grain size and provisioning and scaling mechanisms. Regarding this last aspect, the first prototype of the BECloud tool allows end users to use their own provisioning and scaling tools, to use tools provided by cloud providers or to use the AutoMAP solution [22].…”

“…As a result, the provisioning or scaling mechanisms need to increase the number of virtual machines provisioned at each application tier to fulfil the SLA. Two different provisioning or scaling solutions have been tested, Amazon AutoScaling, [27] and AutoMAP, [22]. In the first case, the scaling solution has been configured to trigger a scale out operation when the average CPU utilization of all virtual machines allocated to a specific application tier is above 90%.…”

Solving recursive equations in analytical modeling of distributed systems

“…These information and messages are usually transmitted in format of packets which are sent directly or using some other system components and with several steps to the desired destination. These conditions can be seen in evaluation of mean time of message delay in interconnection networks, throughput of packets in network on chips, availability of each server in distributed information systems, reliability of a SAAS in a cloud system and also in multitude other similar applications [7][8][9][10][11][12]. In such models, each component of system has constraints in size and hold time of packets for sending them to the next component.…”