+ All Categories
Home > Science > Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

Date post: 07-Aug-2015
Category:
Upload: florian-wessling
View: 108 times
Download: 0 times
Share this document with a friend
Popular Tags:
24
© paluno Towards Exploiting the Full Adaptation Potential of Cloud Applications Clarissa Cassales Marquezan, Florian Wessling, Andreas Metzger, Klaus Pohl PESOS Workshop, ICSE 2014 – Hyderabad, India Chris Woods Karl Wallbom CloudMore, Sweden Intel, Ireland , University of Duisburg-Essen, Germany
Transcript
Page 1: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Towards Exploiting the Full Adaptation Potential of Cloud Applications

Clarissa Cassales Marquezan, Florian Wessling, Andreas Metzger, Klaus Pohl

PESOS Workshop, ICSE 2014 – Hyderabad, India

Chris Woods

Karl WallbomCloudMore, Sweden

Intel, Ireland

, University of Duisburg-Essen, Germany

Page 2: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Structure

2

Motivation and Goals

Conceptual Model

Summary and Future Work

Page 3: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Context

3

CloudWave Framework – High Level Architecture

Page 4: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

4

Physical Nodes

VMs

Applications

Page 5: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

5

Physical Nodes

VMs

Applications

Page 6: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

6

Physical Nodes

VMs

Applications

Physical Nodes

VMs

Applications

Page 7: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

7

Physical Nodes

VMs

Applications

Physical Nodes

VMs

Applications

A

Page 8: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

8

Physical Nodes

VMs

Applications

Physical Nodes

VMs

Applications

A

B

Page 9: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Environment

9

Physical Nodes

VMs

Applications

Physical Nodes

VMs

Applications

A

B

!

Interferences

!

Page 10: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Layers

10

influences

influences

influencesinfluences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

Elements that compose a cloud platform, including Virtualized Resources and their Management

Elements supporting the execution of applications and indicating the Logical Architecture

Elements associated with the Business Logic Implementation

Physical Resources and their Management in a Data Center

Page 11: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Cloud Layers

11

influences

influences

influencesinfluences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

Sources • Literature Survey

• Commercial Solutions

• Industrial Partners

Page 12: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

12

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Page 13: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model – Example

13

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

Page 14: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

14

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

influences

influences

influences

influences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Page 15: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

15

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

influences

influences

influences

influences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Page 16: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

16

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

influences

influences

influences

influences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Page 17: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

17

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

influences

influences

influences

influences

influences

influences

influences

PhysicalLayer

VirtualizationLayer

Logical ApplicationArchitecture Layer

Application BusinessLogic Layer

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

nh

os

ts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Page 18: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

18

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

Page 19: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

19

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

!

A

B

Page 20: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

20

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

!

A

B

Page 21: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

21

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

!

A

B

Page 22: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Conceptual Model

22

<connects

Storage

...

Switch

Router

Virtual Storage

Virtualization Layer

Logical Application Architecture Layer

Application Business Logic Layer

req

ue

sts

connects

Physical Layer

changes

...

ReflectionRe-binding

Mismatch ResolutionDynamic Weaving

Self-adaptation

Management

...

(Re)Deployment

Load Balancing

...

The information necessary to deploy

and run the core logic implementation

is stored in the Configuration entity.

The association between

containers and servers

hosting them creates the

Logical Architecture

vs. Business Logic Mapping

The association between

virtual machine and the

guest OS creates the

Virtualization vs.

Logical Architecture Mapping

The association between

physical and virtual

resources creates the

the Physical vs. Virtual

Mapping.

1 n

changes

Virtual NIC

Logical Architecture

vs. Business Logic

Mapping

req

ue

sts

Virtualization vs.

Logical Architecture

Mapping

Physical vs. Virtualization

Mapping

Load Balancing

...

Load Balancing

Virtual Resource

Embedding

changes

< connects1

n

ho

sts

1

1

hosts

1

n

ho

sts

Configuration

1 n

hosts

Network

Device

NICMemory

CPUDisk

Physical

Machine

Physical Resource

Management

Physical

Resource

AOP

...SOA

OOP

Core Logic

Implementation

Server

Management

...Database

Process

Engine

Application

Server

Web Server

ServerOS Management

Guest OS

Virtual Memory

Virtual CPUVirtual Disk

Virtual Link

Virtual Machine

Resource

Consolidation

Auto-Scaling

Virtual Resource

Management

Virtual

Resource

Container

connects

connects

connects

connects

changes

changes

1

Amazon EC2 Instances

AZ-1

AZ-2

Web Server Web Server

A u t o S c a l i n g G r o u p

App Server App Server

A p p T i e r

CACHE

CACHE

ElastiCacheTier

Amazon RDSM S

Elastic LoadBalancer

Amazon S3Bucket

A

B

Page 23: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Summary and Future Work

23

• CloudWave

• Coordinated Adaptation

• Entities + Adaptation Actions influence each other

• Conceptual Model

• Identify key elements related to adaptation

• Support Developers + Operators

• Future Work

• Tool for Developers

• Adaptation Engine

• Prepare Cloud Monitoring for Coordinated Adaptation (submitted CNSM-Paper)

Page 24: Towards Exploiting the Full Adaptation Potential of Cloud Applications (PESOS Workshop @ ICSE 2014)

© paluno

Summary and Future Work

24

The research leading to these results has received funding from the European Union's Seventh Framework Programme (FP7/2007-2013)

under grant agreement 610802 (CloudWave).

Thank you!


Recommended