Proactive Learning from SLA Violation in Cloud Service based

Application

Ameni Meskini

, Yehia Taher

, Amal El gammal

, Béatrice Finance

and Yahya Slimani

University of Carthage, INSAT, LISI Research Laboratory, Tunis, Tunisia

Laboratoire PRiSM, Universite de Versailles/Saint-Quentin-en-Yvelines, Versailles, France

Faculty of Computers and Information, Cairo University, Cairo, Egypt

Keywords: Cloud Service based Application, SLA Violations Prevention, Cloud Environments, Decision Tree.

Abstract: In recent years, business process management and Service-based applications have been an active area of

research from both the academic and industrial communities. The emergence of revolutionary ICT

technologies such as Internet-of-Things (IoT) and cloud computing has led to a paradigm shift that opens

new opportunities for consumers, businesses, cities and governments; however, this significantly increases

the complexity of such systems and in particular the engineering of Cloud Service-Based Application

(CSBA). A crucial dimension in industrial practice is the non-functional service aspects, which are related

to Quality-of-Service (QoS) aspects. Service Level Agreements (SLAs) define quantitative QoS

objectivesandis a part of a contract between the service provider and the service consumer. Although

significant work exists on how SLA may be specified, monitored and enforced, few efforts have considered

the problem of SLA monitoring in the context of Cloud Service-Based Application (CSBA), which caters

for tailoring of services using a mixture of Software-as-a-Service (SaaS), Platform-as-a-Service (PaaS) and

Infrastructure-as-a-Service (IaaS) solutions. With a preventive focus, the main contribution of this paper is a

novel learning and prediction approach for SLA violations, which generates models that are capable of

proactively predicting upcoming SLAs violations, and suggesting recovery actions to react to such SLA

violations before their occurrence. A prototype has been developed as a Proof-Of-Concept (POC) to

ascertain the feasibility and applicability of the proposed approach.

1 INTRODUCTION

In CSBA (Cloud Service Based Application), a client

can rent a cloud service or a set of cloud services from

a single or multiple providers to create his/her

application. The provisioning of these services relies

on a Service Level Agreement (SLA) (Peter et al.,

2011). In cloud computing, (Brandic et al., 2009)

during the application execution, various events are

produced by several layers (i.e., Cloud and SOA

specific), leading to potential Service Level Objective

(SLO) violations.This crucial dimension in industrial

practice concerns the non-functional service aspects,

which are related to Quality-of-Service (QoS).

Service Level Agreements (SLAs) (Boniface et al.,

2007) define quantitative QoS objectives, which

represents a service contract between the service

provider and the service consumer. The service

provider promises to deliver the requested service

complying with some measurable QoS

objectives/constraints. Typically, a SLA comprises of

a set of Service Level Objectives (SLOs), each of

which represents a quality constraint on the system.

SLAs usually define penalties, in monetary terms that

the provider has to grant to the customer if a QoS is

violated. Furthermore, service-based business

processes have to comply with an ever-growing

number of laws, regulations and standards, such as

Sarbanes-Oxley, Basel-III and ISO 9001. Maintaining

the promised QoS level (George et al., 2003) further

increases the complexity of such systems. The high

dynamism of such systems, and the unprecedented

complexity that arises from the mass of information

that is associated with runtime, puts an emphasis on

their adaptive capabilities. In practice, the rapid

evolving nature of the business and the compliance

domains requires these systems to be equipped with

self-adaptive capabilities to ensure the proper

execution of Cloud Services-Based Applications.

These systems have to autonomously adapt

themselves to changes on service provisioning,

availability of things and content, computing

186

Meskini, A., Taher, Y., gammal, A., Finance, B. and Slimani, Y.

Proactive Learning from SLA Violation in Cloud Service based Application.

In Proceedings of the 6th International Conference on Cloud Computing and Services Science (CLOSER 2016) - Volume 1, pages 186-193

ISBN: 978-989-758-182-3

resources, and network connectivity, while continually

meeting QoS and regulatory constraints. Current

solutions for adaptive Web services and adaptive

service-based business processes fall short to support

essential characteristics of such applications. The

highly rapid and unforeseen adaptive nature of these

applications with their complex and distributed nature

may indirectly affect various layers and components of

the cloud stack (i.e., SaaS, PaaS and IaaS). This puts a

significant emphasis on the monitoring of SLAs, as

well as detecting, predicting and resisting to violations,

which appear to be more challenging in the case of

CSBAs as various providers are dynamically involved.

In the literature, there are two main approaches for

systems monitoring, i.e., reactive and proactive. The

reactive approach mainly reacts after an occurrence of a

violation has occurred. While the proactive approach is

more preventive, by possibly predicting potential future

violations before their occurrence, and by reacting to

avoid their occurrence.

The contributions of this paper are twofold: (i)

we present a novel monitoring framework for

CSBA, namely Proactive learning from SLA

violation, based on MAPE-K

adaptation loop, and

(ii) we concretely address the ‘Analysis’ component

of the proposed framework. This novel proactive

learning approach takes advantage of the massive

amount of past process execution data in order to

predict potential violations. It identifies the best

counter measures that need to be applied. As a

proof-of-concept of the proposed approach, a

prototype has been developed that ascertains the

applicability and feasibility of the proposed solution.

The rest of this paper is structured as follows.

Section 2 discusses related work. Section 3 introduces

a running scenario that will be used throughout the

paper. Section 4 lays the background needed to

understand the work proposed in this paper. The

proposed predictive monitoring framework is

presented in Section 5. Section 6 presents our

proposed SLA violations learning approach. Section

7discusses the implementation of the proposed

approach on a real-life log. Finally, Section 8 draws

conclusions and perspectives.

2 LITERATURE REVIEW

There is a massive amount of work in the literature

related to cloud-based environment, covering

MAPE-K (Monitoring, Analysis, Planning Execution-

Knowledge).

various aspects of this multi-disciplinary domain. In

the next discussion, we are focusing on prominent

efforts in the area of SLA monitoring and

management in CBSA, which is appraised against

the work proposed in this paper.and.

In principle, approaches for monitoring and

detecting SLA violations with respect to QoS

constraints are mainly based on techniques and

strategies to adapt QoS settings according to changes

and violations detected during execution of CSBA.

In this case QoS parameters are generally used to

repair and optimize a web service. Generally, these

adaptive approaches are based on the ability to select

and replace the failed services dynamically at

runtime or during deployment. The selection is

governed not only by the need to substitute services

but to optimize the requirements of QoS of the

system. Accordingly, the system must autonomously

adapt itself in order to improve the quality of service

of the process. In (Tao et al., 2014) proposed a novel

hybrid and adaptive multi learners approach for

online QoS modeling in the cloud; they described an

adaptive solution that dynamically selects the best

learning algorithm for prediction (Leitner et al.,

2011). The proposals in (Fugini et al. 2010) and

(Schmieders et al., 2011) address the problem of

violation detection and adaptation of SLA contracts

between several layers. For example, (Fugini et al.,

2010) proposed a methodology to create, monitor

and adapt the inter-layer SLA contracts. The SLA

model includes parameters such as KPI (key

performance indicators), KGI (indicators key

objectives), and metrics infrastructure. (Schmieders et

al., 2011) proposed a solution to avoid SLA violations

by applying inter-layer techniques. The proposed

approach uses three layers for the prediction of SLA

violations. The identification of adaptation needs is

based on the prediction of QoS, which uses

assumptions about the characteristics of the execution

context. In (Vincent et al., 2015), the authors

introduced a Cloud Application and SLA monitoring

architecture, and proposed two methods for

determining the frequency these applications need to

monitor, they also identified the challenges in regard

with application provisioning and SLA enforcement,

especially in multi-tenant Cloud services.

Discussion: The main limitation of the

aforementioned approaches is that they only

consider certain services regions (execution points)

of the composition and do not consider all process

tasks. Most of the works targeting SLA violations

prediction is addressing grid environments or

service-oriented infrastructure that differs from

cloud infrastructure, therefore the applicability of

Proactive Learning from SLA Violation in Cloud Service based Application

187

these approaches on CSBA is limited. To the best of

our knowledge, the approach proposed in this paper

is the first that uses data mining techniques (Han et

al., 2011) to learn from SLA violations (Vaitheki et

al., 2014) in order to correlate between multiple

violated SLAs. It recommends actions for

automatically reconfiguring the CSBA to avoid the

predicted violation before its occurrence.

3 MOTIVATION SCENARIO

To illustrate the ideas presented in this paper we will

use a simple travel agency scenario, which is

composed of three services: (i) Reserve Flight, (ii)

Payment Service, and (iii) Reserve Hotel Service.

Figure 1: A simple process describing the travel agency

scenario.

Table 1: QoS constraints of SLA relevant to the scenario.

SLA Parameter Value

Response time ≤ 20 Sec

Storage (St) ≥ 2GB

CPU number ≥ 3

Figure 2: Travel agency SLA Violations.

We summarize in Table 1 the Service Level

Objectives (SLOs) for our CSBA. It corresponds to

the SLA specifications of all the QoS constraints for

the whole application. Each cloud service provider

involved in the Travel-Agency-CSBA configuration

promises to satisfy the stipulated Qualities of

Services (QoS) through a Service Level Agreement

(SLA) with his consumer. Each of these services is

made up of a mixture of rented Cloud Services

(SaaS, PaaS and IaaS).This work aims at locating

the failure event and determine adaptation actions in

order to prevent its spread at the others layers, as

soon as possible

The central focus of Travel-

Agency CSBA scenario is the SLA between the

client Travel-Agency and the cloud service

providers offering the Reserve Fly, the Reserve hotel

and the payment cloud services. Upon receiving a

request placed by the customer ‘C’, a process

instance is created. For this instance, the process

execution starts with the activity ‘Reserve Fly’ (S1).

Then, the SLA monitor is called and the software

services are invoked if they are available. The

maximum duration of the Response time of the

whole process should be less than 20 seconds, a

violation of the respective SLA occurs, as it can be

seen in Figure 2.

4 BACKGROUND

Numerous tasks are reached by data mining. They

can be classified in descriptive tasks which are the

association rules in our case and predictive tasks

which is here the decision Tree used for the

prediction from execution logs. In this section, we

first introduce the decision Tree, a commonly used

data mining technique in order to build predictions

models from execution logs to be able to predict

potential violation and react to it proactively. This is

followed by an overview of association rules.

4.1 Decision Tree

Objective: classification of people or things into

groups by recognizing patterns. The user or the

expert has always a tendency to structure or classify

data into groups of similar objects called classes. For

this purpose, he uses distance measurements in order

to evaluate the belonging of an object to a class. The

most known classification methods are nearest

neighbor and decision trees. A decision tree seeks to

represent the studied objects in a tree, according to a

hierarchy of attributes. Decision trees are popular

because they provide a synthetic representation of

data. They are graphical representations of a

classification procedure aiming to derive a result

from a test of attributes (internal nodes of the tree).

A node represents a class, an arc represents a

partitioning predicate of the source class and the

leaves of the tree are the classes we want to predict

or explain their statements. CART (Breiman et al.,

1984) and C4.5 are among the best known

algorithms for generating decision tree. These

algorithms generate classification rules easy to

interpret by the user. The generated rules can be

used to build predictive models. A decision tree is

CLOSER 2016 - 6th International Conference on Cloud Computing and Services Science

188

used to classify records by hierarchical division into

subclasses.

4.2 Association Rules

Objective: associating what events are likely to

occur together.

Association models aim to discover relationships

or correlations in a set of items. Association rules is

a data mining technique intending to find associated

values in a given dataset and serving decision

making. It has the following form: A->B (S %), (C

%). This rule means that tuples satisfying conditions

in A also satisfy conditions in B.

A rule is always given with two measures (the

support (S) and the confidence (C)) describing its

strength and its interestingness. The support

(equation (1)) is the percentage of transactions that

satisfy A and B among all the transactions of the

transactions base. The confidence (equation (2)) is

the percentage of transactions that verify the

consequent of a rule among those that satisfy the

antecedent (premise) data.A rule is always given

with two measures (the support (S) and the

confidence (C)) describing its strength and its

interestingness. The support (equation (1)) is the

percentage of transactions that satisfy A and B

among all the transactions of the transactions base.

The confidence (equation (2)) is the percentage of

transactions that verify the consequent of a rule

among those that satisfy the antecedent (premise)

data.





→



∪)

(1)

Confidence A → B = P (B/A) = Support

AUB/Support (A)

(2)

The extraction of association rules is the generation

of the interesting rules with support and confidence

greater than minimum thresholds of support and

confidence. The process of extracting association

rules involves two distinct phases. Firstly, the items

having a support level that exceeds a certain

threshold are segregated. Secondly, the most

frequent items are combined in order to generate

associations (Chelghoum, 2004).

5 A FRAMEWORK FOR

PROACTIVE LEARNING

FROM SLA VIOLATIONS

Figure 3 portrays a high-level architectural view of

the proposed cross-layer self-adaptation framework.

The framework is based on MAPE-K adaptation

loop, introduced by IBM as an efficient and novel

approach for self-adaptation in autonomic

computing. As shown in Figure 3, the MAPE-K

adaptation loop comprises of five main components

corresponding to its acronym, which will be

discussed in the following. The main focus of this

paper is on the Analysis component of the MAPE-K

loop, where a proactive learning approach is

proposed (cf. Section 5) to predict potential QoS

violations based on historical execution logs and

react accordingly to avoid/prevent the predicted

violation. Therefore, the upcoming sections will

focus on presenting the details of these two main

components.

Figure 3: High-level architectural view of the proposed

proactive monitoring framework.

Knowledge: SLA is a predominant entity in cloud

service based systems (CSBS). In CSBS, clients rent

services from providers instead of buying services.

This means that a CSBS is a composition of a list of

rented services. Thus, SLA becomes critical to both

service clients and providers and it needs to be

monitored constantly while a CSBS is running with

the aim to not only detect violations but also prevent

them. As in CSBS a number of providers are

involved, detecting and resisting violations (of

multiple SLAs that engage different providers form

different locations) are enormously challenging.

Therefore, an efficient and effective approach is of a

paramount importance for CSBS. As CSBS rely on

third party cloud service providers, an SLA involves

a ‘consumer’ and one to many ‘providers’.

Monitoring: Monitoring of SLA compliance is of

crucial importance to the proposed framework.

Monitoring is intended to be in a near real-time

fashion in order to take corrective actions before it is

too late (Yehia et al. 2014). Our intention also is to

Proactive Learning from SLA Violation in Cloud Service based Application

189

predict possible SLA violation and avoid them

before they occur. To tackle the monitoring task, we

will depend on complex event processing (CEP)

technology.

Analysis: Analysis is based on the monitoring

results; the analysis component is responsible for

performing complex data analysis and reasoning, by

the continuous interaction with the knowledge

component. Based on information extracted from the

historical traces, predictions and recommendations

are provided for running instances. Such predictions

and recommendation rely on data mining techniques,

more specifically, decision trees.

Planning: Once a violation is predicted, the planning

component takes the hand over. To deal with such a

predicted violation, the planning component starts by

searching for alternative solutions in order to avoid

the occurrence of the predicted violation. The

planning component will attempt to adapt the smallest

possible set of services without directly targeting a re-

engineering process of the whole system.

Execution: Based on the results of the prediction

models constructed in the previous planning

component, the execution component is responsible

for selecting the adaptation plan (in the form of

recommendations as passed from the planning

component) with the highest probability of success.

This evaluation will iteratively enhance the quality

of the predication models by better learning (the

feedback loop in Figure 3).Our work in this

component is ongoing.

6 THE PROPOSED LEARNING

APPROACH IN CLOUD

ENVIRONMENTS

In this section, we present the details of the proposed

learning approach, which combines different existing

techniques ranging from learning approaches to

decision tree learning, to provide predictions, at

runtime, about the achievement of business goals in a

Business Process (BP) execution trace. In the

following sections, we provide an overview of the

approach. Section 7 discusses the implementation of

the proposed approach as a Proof-Of-Concept (POC)

of the realization of the proposed approach.

6.1 The Proposed Learning Approach

The process of learning from SLA violation and

making dependencies precedence between different

SLAs violations in CSBAs have been identified as

major research challenges in Cloud environments.

Figure 4: Proposed SLA Violation Learning approach:

Architecture overview.

SLA does not contain information about the

dynamicity of the system. In other words, it is

independent of the context of the business process,

and it contains information about the service

behavior or quality provided by the service which

we aim to exploit. SLAs are not mathematically

defined. That means that the semantics of the SLA

elements and metrics are defined in natural

languages, which makes it harder to understand the

semantic of QoS, and it is usually dependent on the

client and provider contract. Thus, being precise and

formal about SLA semantic is necessary. SLAs

violations come from different kind of failures,

determining the appropriate type of actions to be

taken when predicting an SLA violation is equally

important.

First Phase: Learning phase: It is a continuous

evolving process. The association rules extraction is

explained as follows:

Given: a set of historical BP event logs of SLA

violations.

Find: Association rules

Figure 5: Learning Phase.

Our method of Association Rules (AR)

determination goes through three steps:

CLOSER 2016 - 6th International Conference on Cloud Computing and Services Science

190

 The first step is to discover frequent itemsets.

 The second step is dedicated to find out ARs on

the basis of the first step outputs.

 The third step is the refinement of the extracted

ARs.

These steps are depicted in the diagram shown in

Figure 5. Two steps are combined in order to carry

out ARs. They are respectively the generation of

frequent itemsets and the extraction of association

rules. The frequent itemsets are extracted as defined

in Algorithm 1.

Algorithm 1: Extraction and refinement of AR.

Inputs: Execution logs

Outputs: ARs

Reject-rule: = Boolean

Begin

Reject-rule: =False

(1) Find all the frequent itemsets

by C4.5 execution (execution

logs)

(2) Find all the possible ARs by

C4.5 execution

(3) Save the obtained ARs

End

(4) Save the rules set containing

only correct rules

End

The input corresponds to the set of historical data

generated by the previous SLA violation of CSBA.

The AR contains QoS property of SLA in the

antecedent and the violated SLA as consequent of

the rule. The proposed formula of the AR is given in

equation:

eventcondition



perform



action

The next phase is the prediction phase (described

below), in that phase some historical executions of

CSBA are necessary to bootstrap the prediction. The

concrete amount of instances that are necessary,

depend both on the expected quality of prediction

and on the size and complexity of the service

composition.

Second Phase: SLA Prediction

The objective of this phase is to (i) predict potential

violation, and to (ii) construct/build the best

configuration as the recommendation from what

have been detected and learned based on the

previous learning phase.The set of such entries is

presented in the Decision Tree of the Figurewhere

the output is the frequent set of dependencies.The

Prediction algorithm gives precise predictions and

avoids unnecessary adaptations. Generally, the

approach that predicts SLA violations is based on

the idea of predicting concrete SLO values based on

whatever monitoring information is already

available. In order to identify which data should be

used to train which model, some domain knowledge

is necessary. However, dependency analysis can be

used to identify the factors that have the biggest

influence on the respective SLOs.

The association rules prediction is explained as

follows:

Given: ARs extracted.

Find: Predictive ARs.

The process of this phase is shown in Algorithm 2:

Algorithm 2: Prediction of predictive AR.

Inputs: ARs

Outputs: Predicted ARs

Reject-rule: = Boolean

Begin

Reject-rule: =False

(5) Compare all the obtained ARs in

different time intervals

(6) Suggest predicted Rules

End

(7) Save the predicted rules set

containing only correct rules

End

6.2 Exemplifying on the Running

Scenario

In this section, the prediction of violations is applied

on the Travel agency CSBA scenario (described in

section 3). The rules below are an outcome of the

Decision Tree mining the data sets sent as inputs

from an Excel file of the Travel-Agency scenarios is

described in (section 3).

Example of rule

IF sum RT2+RT3 ≤ 10 = no And RT3≤ 6AndRT2 ≤

11 And RT1≥ 5 Then Violation = Viol (P1, P6), the

configuration will be: {Response time ≤ 20, 1CPU,

2GB RAM}.

As presented in Figure 2, IaaS is an

infrastructure as a service that is a rented service

from amazon service provider with 1 CPU and 2 GB

Ram and it promises PaaS is a platform as a service

that presents a rented IIS (internet information

services), with 1 CPU and 2 GB Ram and it

promises to satisfy response time <20 sec. The

Global SLA service promises to satisfy response

time <20 sec, giving 10 sec as a response time for

Proactive Learning from SLA Violation in Cloud Service based Application

191

Table 2: Proactive Actions.

Violation ID Violation Type Action Type Action

Viol1 Violation Qos Reparation Raise the RAM capacity to 2GB

Viol 2 Availability Substitution Change violated service

Viol 3 Security Reparation Adapt the service to security police

each of ‘Reserve Hotel’ and 5 sec for ‘Reserve

Flight’, and 5 sec for ‘Payment Application’ service.

The response time for example could be violated

when any of these application services has an

internal violation in response time at the level of

SaaS, PaaS or IaaS. In order to avoid such situation,

the SLA manager acts proactively based on history

of completed activities. At the same time, another

monitoring component detects that there is an I/O

failure at the SaaS layer as S1 has produced a wrong

output. A specific rule is triggered that derives the

best strategy which consists of executing another

instance of the web service on a more powerful

server with a better memory and CPU allocation.

(Amazon (3cpu, 3GB RAM)). Assume that a

Monitoring Component, running at the server where

the web service is executed, detects that the available

main memory is not sufficient (IaaS layer) for the

web service. At this stage, proactive actions are

suggestions to be taken based on some predefined

suitable actions for each type of violation the system

may encounter. In Table 2, we can see that each

violation has a violation type that could be

availability or security depending on the type of

violation. The actions taken are of two types, namely

surgery and elastic actions.

7 IMPLEMENTATION

To demonstrate the applicability and feasibility of

our approach, we developed a prototype

using

JAVA. We trigger the execution of 100 process

instances using a test client. For each of these

instances we select the concrete supplier service and

shipper service randomly in order to ensure that

history data used for learning contains metrics data

on each of these services. During process instance

execution, the previously specified metrics are

measured and saved in the knowledge database.

Then, for each checkpoint a decision tree is learned

A video demonstration is available at: https://www.youtube.

com/watch?v=oDEFYGBPdH0

using the J48 algorithm. For the implementation of

the Predictor, we rely on the WeKa J48

implementation of the C4.5 algorithm, which takes

as input a‘.arff’ file and builds a decision tree as

shown in Figure 6: Text files of real data. The ‘.arff’

file contains a list of typed variables (including the

target variable) and, for each trace prefix (e.g., for

each data snapshot), the corresponding values are

also maintained. The resulting Decision Tree is then

analyzed to generate predictions and

recommendations as shown in Figure 6.

The

configuration manager is responsible for configuring

the CSBA. The proactive actions suggested by the

proactive engine are mapped into the configuration

manager to take the action. The action taken is then

stored in the knowledgebase. The algorithm searches

in the database (as shown in (Figure 7) for suitable

actions that can be used. Below in figure 8 is a part

of the Excel file that we used for our decision. For

example, as shown in the Table 3 since the violation

is Response Time, then the suitable action is to add 1

CPU to the violating service.

Figure 6: Text files of real data.

We evaluated experimentally the model’s

performance and accuracy. The experiments were

performed on a machine with quad-core CPU 2.6

GHz, 8GB RAM and Mac OS X operating system.

This experiment evaluates the algorithm’s raw

CLOSER 2016 - 6th International Conference on Cloud Computing and Services Science

192

relevant and absolute accuracy. The static metrics

precision and recall is measured while fluctuating

the interval size from 4 to 20 events. Figure 8 shows

that relevant precision is 1 for small intervals and

falls while increasing the interval size, while

absolute precision fluctuates similarly at lower

levels (as more irrelevant sub-patterns are

discovered).

Figure 7: A screenshot of ARs Extraction.

Figure 8: Evaluation results.

8 CONCLUSION AND FUTURE

WORK

In this paper, we proposed a proactive framework

for learning from SLA violations in Cloud Service

Based application. We proposed a proactive

approach that uses business process execution logs

to learn from past violations and extracts knowledge

between violated SLA’s, This is based on building

predictions models that is capable of predicting

potential violations before their occurrence, as well

as recommending proactive response

recovery/preventive actions. As future work

directions, we plan to design and implement the

planning and Execution components of the

comprehensive framework for automatic cross-layer

self-adaptation of service-based business processes

running on cloud environments, proposed in this

paper. Future work will also focus on validating the

approach by applying it to large-scale case studies

from diverse problem domains.

REFERENCES

Boniface M., PhillipsS. C., A. Sanchez-Macian, and

Surridge M.. Dynamic service provisioning using

GRIA SLAs. In ICSOC’07 workshops, 2007.

Brandic, I.,Towards self-manageable cloud services. In

33rd Annual IEEE COMPSAC’09, 2009.

Breiman, L., Friedman, J.H., Olshen, R.A., Stone, and C.J.

(1984). Classification andRegression Trees. Edition

Wadsworth & Brooks.Montery, California, ISBN 0-

412-04841-8, 358 pages.

Chelghoum, N., (2004). Fouille de données spatiales, Un

problème de fouille de donnéesmulti-tables. Thèse de

doctorat présentée et soutenue publiquement à

l’université deVersailles Saint-Quentin-en-Yvelines

U.F.R de sciences Par Nadjim CHELGHOUM le16

décembre 2004.

Fugini,M., Siadat, H.,SLA contract for Cross-Layer

monitoring and adaptation. In S. Rinderle-Ma, S.

Sadiq, and F. Leymann, editors, BPM Workshops,

pages 412--423., 2010.

Han, J., Kamber, M., Pei, J., , Data Mining: Concepts and

Techniques. (3rd ed.)Morgan Kaufmann, USA (2011).

Leitner, P., Michlmayr, A., Rosenberg, F., Dustdar, S.,

Monitoring, Prediction and Prevention of SLA

Violations in Composite Services, icws, pp.369-376,

2010 IEEE SCC, 2010.

Peter, M., Timoth, y G., "The NIST definition of cloud

computing," 2011.

Schmieders, E., Micsik, A., Oriol, M.,Mahbub, K., and

KazhamiakinR., Combining SLA prediction and cross

layer adaptation for preventing SLA violations”. In

Proceedings of the 2nd Workshop on Software

Services: Cloud Computing and Applications based on

Software Services, Timisoara, Romania, June 2011.

Tao, C., Rami, B., Xin, Y., Online QoS Modeling in the

Cloud: A Hybrid and Adaptive Multi-Learners Approach,

the 7th IEEE/ACM UCC, London, UK, 2014.

Vaitheki, K.,Urmela, S., A SLA violation reduction

technique in Cloud by Resource Rescheduling

Algorithm (RRA), International Journal of Computer

Application and Engineering Technology Volume 3-

Issue 3, July 2014. Pp. 217-224.

Vincent, C.,Emeakaroha, Marco, A., S.,Netto,

IvonaBrandic, César, A., F., De Rose, Application-

Level Monitoring and SLA Violation Detection for

Multi-Tenant Cloud Services, Emerging Research in

Cloud Distributed Computing Systems, 2015.

Yehia, T., Rafiqul, H., Dinh Khoa, N., Béatrice, F.,

:PAEAN4CLOUD: A Framework for Monitoring and

Managing the SLA Violation of Cloud Service-based

Applications. CLOSER 2014: 361-371.

Proactive Learning from SLA Violation in Cloud Service based Application

193