Transcript
Page 1: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Daniel Moldovan,Georgiana Copil, Hong-Linh Truong,

Schahram Dustdar

MELA: Monitoring and Analyzing Elasticity of Cloud Services

Work partially supported by the European Commission in terms of the CELAR FP7 project (http://www.celarcloud.eu/)

Distributed Systems Group (http://dsg.tuwien.ac.at/)

Vienna University of Technology (http://www.tuwien.ac.at/)

Page 2: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

MotivationElastic Cloud Service

2

Data-as-a-Service for Machine to Machine platforms Load balancer distributes incoming requests to Event Processing instances Distributed Data Store: Controller and Nodes

Start with an initial lighter configuration

Page 3: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

MotivationElastic Cloud Service

2

Data-as-a-Service for Machine to Machine platforms Load balancer distributes incoming requests to Event Processing instances Distributed Data Store: Controller and NodesAdd service unit instance when load

increases

Page 4: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

MotivationElastic Cloud Service

2

Data-as-a-Service for Machine to Machine platforms Load balancer distributes incoming requests to Event Processing instances Distributed Data Store: Controller and NodesRemove service unit instance when load

decreases

Page 5: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

MotivationElastic Cloud Service

2

Data-as-a-Service for Machine to Machine platforms Load balancer distributes incoming requests to Event Processing instances Distributed Data Store: Controller and Nodes

Add service unit instance and data node instance when load increases too much

Page 6: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Service Level Monitoring Response time Number of clients Other specific metrics

System Level Monitoring Ganglia, Nagios, etc. CPU usage Memory usage Network transfer

User-Defined Requirements violation: - Cost per client too highReasons: - Too much logging? Monitoring chatter? - Too expensive VMs? Which one can be downsized? - Not enough clients? Why?

Controlling the service’s elasticity

3

MotivationInsufficient Cloud Service Monitoring and Analysis Support

Page 7: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Approach and ChallengesStructure Monitoring Data

How to map system data to service level? How to derive higher level information?

4

Monitoring Data

Service Structure

Impose service structure over collected monitoring data

Page 8: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Metrics composition and enrichment

Page 9: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Page 10: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Enrich metric with COST information

COST/VM * numberOfVMs

Page 11: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Propagate activeConnections from LoadBalancer service unit

Page 12: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Page 13: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Page 14: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Monitoring Snapshot

5

Compute cost/client/h

Page 15: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Evaluate Service’s Elasticity How to characterize service elasticity? How to derive service‘s behavior limits? How to characterize and predict elasticity behavior?

Approach and Challenges

6

Page 16: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Runtime Properties of Elastic Cloud Services Background

Elastic process: cost, quality and resources elasticity Extend concept to cloud services

Elasticity Space Collection of monitoring snapshots I.e. the space in which an elastic service moves

Elasticity Boundary Elasticity Space boundaries in which service’s requirements are

respected

Elasticity Pathway Characterizes service evolution trough elasticity space

Elasticity Dimensions

16

Page 17: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Elasticity SpaceEvent Processing Topology

8

Elasticity Space Snapshot

Elasticity Space “Clients/h” Dimension

Elasticity Space “Response Time” Dimension

Service requirement COST<= 0.0034$/client/h 2.5$ monthly subscription for each

service client (sensor)

Page 18: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Elasticity SpaceEvent Processing Topology

Service requirement COST<= 0.0034$/client/h 2.5$ monthly subscription for each

service client (sensor)

8

Elasticity Space “Clients/h” Dimension

Elasticity Space “Response Time” Dimension

Determined Elasticity Space Boundaries Clients/h > 148 300ms ≤ ResponseTime ≤ 1100 ms

Page 19: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Elasticity Pathway

9

Service requirement COST<= 0.0034$/client/h 2.5$ monthly subscription for

each service client (sensor)

Page 20: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Elasticity Pathway

9

Cloud Service Elasticity Pathway

Service requirement COST<= 0.0034$/client/h 2.5$ monthly subscription for

each service client (sensor)

Page 21: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Multi-Level Elasticity Pathway

9

Event Processing service unit Elasticity Pathway

Cloud Service Elasticity Pathway

Service requirement COST<= 0.0034$/client/h 2.5$ monthly subscription for

each service client (sensor)

Page 22: MELA: Monitoring and Analyzing Elasticity of Cloud Services -- CloudCom 2013

Conclusions

10

Concepts Elasticity Space and Elasticity Boundary Elasticity Pathway

Mechanisms Constructing cross-layer monitoring snapshots Determining elasticity space and boundary Determining elasticity pathway

MELA Customizable framework for monitoring and

analyzing elasticity of cloud services

MELA: Monitoring and Analyzing Elasticity of Cloud Services

Work partially supported by the European Commission in terms of the

CELAR FP7 project (http://www.celarcloud.eu/)

Distributed Systems Group(http://dsg.tuwien.ac.at/)

Vienna University of Technology (http://www.tuwien.ac.at/)

http://dsg.tuwien.ac.at/research/viecom/mela/


Top Related