Cloud Computing Architectures and Design Issues

Transcription

1 Cloud Computing Architectures and Design Issues Ozalp Babaoglu, Stefano Ferretti, Moreno Marzolla, Fabio Panzieri {babaoglu, sferrett, marzolla,

2 Outline What is Cloud Computing? A View from Bologna Design Issues Energy Reduction P2P Cloud QoS-aware Clouds Seminari DISI 2012 / Cloud Computing 2

3 What is Cloud Computing? Cloud computing is a model for enabling ubiquitous, convenient, on-demand network access to a shared pool of configurable computing resources (e.g., networks, servers, storage, applications, and services) that can be rapidly provisioned and released with minimal management effort or service provider interaction. The NIST Definition of Cloud Computing, Special Publication Seminari DISI 2012 / Cloud Computing 3

4 Main Characteristics On-demand self-service. Broad network access. Resource pooling. Rapid elasticity. Measured service. Seminari DISI 2012 / Cloud Computing 4

5 Service Models Software as a Service (SAAS) Platform as a Service (PAAS) Application Platform Google Apps, SalesForce.com Microsoft Azure, Google AppEngine Infrastructure as a Service (IAAS) Infrastructure Hardware Amazon EC2 Data Center Qi Zhang, Lu Cheng, Raouf Boutaba, Cloud computing: state-of-the-art and research challenges, J Internet Serv Appl (2010) 1: 7 18 Seminari DISI 2012 / Cloud Computing 5

6 Deployment Model Public Cloud Cloud infrastructure available to the general public Community Cloud Cloud infrastructure available to a community of users with shared concerns Private Cloud Cloud infrastructure available to a single organization Hybrid Cloud Any combination of the above Seminari DISI 2012 / Cloud Computing 6

7 Seminari DISI 2012 / Cloud Computing 7

8 Bologna F. Panzieri, O. Babaoglu, V. Ghini, S. Ferretti, M. Marzolla, Distributed Computing in the 21st Century: Some Aspects of Cloud Computing, Volume 6875 of LNCS, October 2011, ISBN Seminari DISI 2012 / Cloud Computing 8

9 A view from Bologna - HPC - Data analytics - MMOG - Private clouds - Mobile gateways - Loosely coupled applications Applications Cloud API SLA mgmt SLA mgmt SLA mgmt QoS policy management Cloud core infrastructure Virtualization Virtualization Virtualization Virtualization Layer Centralized Federated Peer-to-peer Seminari DISI 2012 / Cloud Computing 9

10 A view from Bologna - HPC - Data analytics - MMOG - Private clouds - Mobile gateways - Loosely coupled applications Applications Cloud API SLA mgmt SLA mgmt SLA mgmt QoS policy management Cloud core infrastructure Virtualization Virtualization Virtualization Virtualization Layer Centralized Federated Peer-to-peer Seminari DISI 2012 / Cloud Computing 10

11 Energy Reduction through gossiping - HPC - Data analytics - MMOG - Private clouds - Mobile gateways - Loosely coupled applications Applications Cloud API SLA mgmt SLA mgmt SLA mgmt QoS policy management Cloud core infrastructure Virtualization Virtualization Virtualization Virtualization Layer M. Marzolla, Centralized O. Babaoglu, F. Panzieri. Federated Server Consolidation Peer-to-peer in Clouds through Gossiping. In Proc. of First IEEE Int. Workshop on Sustainable Internet and Internet for Sustainability, Lucca, Italy, June Seminari DISI 2012 / Cloud Computing 11

12 Electricity use breakdown Toward energy efficient computing, ACM Queue, Seminari DISI 2012 / Cloud Computing 12

13 Virtualization Main feature of a Cloud system Dynamic scalability (pay-as-you-go economic model) Virtualization of resources Assumptions Physical resources (servers) are multi-core/multi-processor machines; each server can support up to C VM instances Goal Users request Virtual Machine (VM) instances Users release instances when no longer needed The VM monitor supports live migration of VMs Minimize energy consumption by consolidating VMs Seminari DISI 2012 / Cloud Computing 13

14 VM Consolidation VM 1 VM 2 VM 3 VM 4 VM Monitor OS Host 1 VM Monitor OS Host 2 (a) Before consolidation VM Monitor OS Host 3 VM 1 VM 1 VM 2 VM 3 VM 4 VM 4 VM Monitor VM Monitor VM Monitor OS OS OS Host 1 Host 2 Host 3 (b) After consolidation of VM 1 and VM 4 to host 2 Seminari DISI 2012 / Cloud Computing 14

15 VM consolidation through gossiping Each server hosts the V-MAN daemon Daemons maintain an overlay network such that each daemon is connected to at most K other nodes Daemons exchange messages only with neighbors The overlay is maintained with the Newscast algorithm V V V V V V V V V V V V Seminari DISI 2012 / Cloud Computing 15

16 Example (1) Capacity = 4 VM VM VM Host 1 Host 2 Seminari DISI 2012 / Cloud Computing 16

17 Example (1) Capacity = 4 VM VM VM Host 1 Host 2 I have two VMs Seminari DISI 2012 / Cloud Computing 17

18 Example (1) Capacity = 4 VM VM VM Host 1 Host 2 I send you one Seminari DISI 2012 / Cloud Computing 18

19 Example (2) Capacity = 4 VM VM VM VM VM Host 1 Host 2 Seminari DISI 2012 / Cloud Computing 19

20 Example (2) Capacity = 4 VM VM VM VM VM Host 1 Host 2 I have two VMs Seminari DISI 2012 / Cloud Computing 20

21 Example (2) Capacity = 4 VM VM VM VM VM Host 1 Host 2 Send me one Seminari DISI 2012 / Cloud Computing 21

22 Performance assessment We implemented V-MAN using the cycle-driven simulator engine provided by PeerSim (peersim.sf.net) Parameters: K=20 (each node maintains a list of 20 neighbors) C=8 (maximum capacity of each host is 8 VMs) Topology is managed using Newscast Length of each simulation run is 20 steps Results are averages of 10 independent simulation runs Results: F 0 Fraction of empty hosts F 0, opt Optimal fraction of empty hosts Seminari DISI 2012 / Cloud Computing 22

23 Static System Seminari DISI 2012 / Cloud Computing 23

24 Animation Seminari DISI 2012 / Cloud Computing 24

25 Dynamic system Seminari DISI 2012 / Cloud Computing 25

26 Dynamic System with failures Seminari DISI 2012 / Cloud Computing 26

27 A P2P Cloud - HPC - Data analytics - MMOG - Private clouds - Mobile gateways - Loosely coupled applications Applications Cloud API SLA mgmt SLA mgmt SLA mgmt QoS policy management Cloud core infrastructure Virtualization Virtualization Virtualization Virtualization Layer O. Centralized Babaoglu, M. Marzolla, M. Federated Tamburini. Design and Peer-to-peer Implementation of a P2P Cloud System. To appear in Proc. of the 27th ACM Symposium on Applied Computing (SAC 2012), Trento, Italy, March Seminari DISI 2012 / Cloud Computing 27

28 Self-managing P2P Cloud Assemble a Cloud out of individual devices E.g. low-power devices such as set-top boxes, ADSL modems, New business model to harness the computational power of otherwise idle devices Individual devices leave and join, but the Cloud keeps a coherent structure anyway No central controller Seminari DISI 2012 / Cloud Computing 28

29 Self-managing P2P Cloud O. Babaoglu, M. Jelasity, A-M Kermarrec, A. Montresor, M. van Steen, Managing clouds: a case for a fresh look at large unreliable dynamic networks ACM SIGOPS Operating Systems Review, 2006 Seminari DISI 2012 / Cloud Computing 29

30 P2P Cloud Goals Implement fully decentralized monitoring and Management capabilities Allocate x% of available nodes for a given task Allocate at least n node for a given task How many nodes are currently busy? How many compute hours have been consumed by user X? Seminari DISI 2012 / Cloud Computing 30

31 P2P Cloud Architecture User Node User Interface P2PCS Daemon Node-to-Node Interface Seminari DISI 2012 / Cloud Computing 31

32 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Bootstrapping Service Peer Sampling Service = implemented modules Seminari DISI 2012 / Cloud Computing 32

33 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Gather an initial set of nodes to start the message exchange Bootstrapping Service Peer Sampling Service = implemented modules Seminari DISI 2012 / Cloud Computing 33

34 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Bootstrapping Service Peer Sampling Service = implemented modules Provide each node with a list of peers to exchange messages with Seminari DISI 2012 / Cloud Computing 34

35 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Bootstrapping Service Rank the nodes according to one attribute (e.g., 5% of the total n. of nodes; 1% fastest nodes;...) Peer Sampling Service = implemented modules Seminari DISI 2012 / Cloud Computing 35

36 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Bootstrapping Service Peer Sampling Service Compute global measures (e.g., network size) using local message exchange = implemented modules Seminari DISI 2012 / Cloud Computing 36

37 P2P Cloud Architecture Authentication / Authorization layer Instance Management API Monitoring API Storage API Dispatcher Slicing Service T-Man Monitoring System Aggregation Service Storage System Bootstrapping Service Peer Sampling Service Build an overlay network with a given topology (e.g., tree, ring, mesh...) = implemented modules Seminari DISI 2012 / Cloud Computing 37

38 Aggregation example Computing the mean X Y X X Y X Y Y (X+Y)/2 (X+Y)/2 Seminari DISI 2012 / Cloud Computing 38

39 P2PCS: Building subclouds slice 1 slice slice (a) (b) (c) Seminari DISI 2012 / Cloud Computing 39

40 P2PCS API run-nodes subcloud_id number Creates a subcloud with number nodes; subcloud_id is set as the name of the newly created subcloud terminate-nodes subcloud_id nodename1 nodenamen Removes the named nodes from the subcloud with given id add-new_nodes subcloud_id number Adds number nodes to the subcloud identified by subcloud_id. The new nodes are chosen without any particular criteria describe-instances nodename Prints a human-readable description of the given node monitor-instances Return the global size of the Cloud using the aggregation service unmonitor-instances Stops printing the global size of the Cloud Seminari DISI 2012 / Cloud Computing 40

41 QoS-aware Clouds - HPC - Data analytics - MMOG - Private clouds - Mobile gateways - Loosely coupled applications Applications Cloud API SLA mgmt SLA mgmt SLA mgmt QoS policy management Cloud core infrastructure Virtualization Virtualization Virtualization Virtualization Layer Centralized Federated Peer-to-peer S. Ferretti, V. Ghini, F. Panzieri, M. Pellegrini, E. Turrini, QoS-aware Clouds, in Proc. 3rd Int. Conf. on Cloud Computing (IEEE Cloud 2010), Miami (USA), IEEE, July Seminari DISI 2012 / Cloud Computing 41

42 Motivations QoS: crucial factor for the success of cloud computing providers if not delivered as expected, it may impact the provider s reputation Compliance to SLA SLA: legally binding contract stating the QoS guarantees required by cloud customer typically includes max response time, throughput, error rate may include non functional requirements such as timeliness, scalability, availability we address response time, only Seminari DISI 2012 / Cloud Computing 42

43 SLA Example <ContainerServiceUsage name="highprority requestrate="100/s"> <Operations>... </Operations>... </ContainerServiceUsage> Customer obligations and rights Maximum request rate from the client application to the virtual execution environment in the cloud Operations the application is allowed to invoke Seminari DISI 2012 / Cloud Computing 43

44 SLA Example <ServerResponsabilities serviceavailability="0.99 efficiency="0.95" efficiencyvalidity="2"> <OperationPerformance name="highpriority maxresponsetime="1.0s"> <Operations> </Operations> </OperationPerformance>... </ServerResponsabilities> Responsibilities of a service running in the cloud Seminari DISI 2012 / Cloud Computing 44

45 SLA Example <ServerResponsabilities serviceavailability="0.99 efficiency="0.95" efficiencyvalidity="2"> <OperationPerformance name="highpriority maxresponsetime="1.0s"> <Operations> </Operations> </OperationPerformance>... </ServerResponsabilities> Probability that the service is available over a predefined time period Responsibilities of a service running in the cloud Seminari DISI 2012 / Cloud Computing 45

46 SLA Example <ServerResponsabilities serviceavailability="0.99 efficiency="0.95" efficiencyvalidity="2"> <OperationPerformance name="highpriority maxresponsetime="1.0s"> <Operations> </Operations> </OperationPerformance>... </ServerResponsabilities> Fraction of SLA violations that can be tolerated, within a predefined time interval, before the service provider incurs a penalty Responsibilities of a service running in the cloud Seminari DISI 2012 / Cloud Computing 46

47 SLA Example <ServerResponsabilities serviceavailability="0.99 efficiency="0.95" efficiencyvalidity="2"> <OperationPerformance name="highpriority maxresponsetime="1.0s"> <Operations> </Operations> </OperationPerformance>... </ServerResponsabilities> Required service responsiveness Responsibilities of a service running in the cloud Seminari DISI 2012 / Cloud Computing 47

48 SLA Example <InfrastructureSLA> <node typename="alpha"> <cpu num="4" type="x86-64 performance="x"/> <ram dim="64gb" performance="y"/> <storage dim="640gb performance="z"/> <network> <if name="eth0"> <incoming rate="5mbit" peak="10mbit" maxburst="10mb"/> <outgoing rate... /> </network> </node>... </InfrastructureSLA> Network characteristics Hardware characteristics SLA between the service provider and the cloud infrastructure hosting that service Seminari DISI 2012 / Cloud Computing 48

49 QoS-aware Cloud Architecture Main components Load balancer Monitoring service SLA policy engine Configuration service reconfiguration request Seminari DISI 2012 / Cloud Computing 49

50 Load Balancer Implements the load dispatching and balancing functionalities Receives requests from clients and dispatches them to virtual resources, balancing the load Incorporates a SLA Policy Engine which analyzes logs of the Monitoring Service to identify SLA violations Seminari DISI 2012 / Cloud Computing 50

51 Monitoring Service Monitors the environment to detect QoS deviations from what specified in the SLA The component within the Load Balancer monitors incoming requests and related responses There is a Monitoring Service instance for each virtual resource Seminari DISI 2012 / Cloud Computing 51

52 Configuration Service Responsible for both configuration and run-time reconfiguration of the application hosting environment Dynamically resizes the resources for the service by adding/removing them as needed Seminari DISI 2012 / Cloud Computing 52

53 Experimental Evaluation Request Generator Load Balancer Response Generator Monitoring System SLA Policy Enginer Configuration Service Tool implements principal components Load balancing policies QoS handling policies Assumptions: required hosting SLA efficency = 95%, VM allocation time = 2s Other tests carried out with VM allocation time = 6s, 10s VM allocation can take up to 400s Seminari DISI 2012 / Cloud Computing 53

54 Preliminary Results Load progressively increased until reaching 90 requests per sec, then pogressively decreased a) Response time VMs allocated as load increases and released as it decreases a) Violation Rate Peaks occur a new VM is added Seminari DISI 2012 / Cloud Computing 54

55 Preliminary Results Augmented load till reaching 13 VMs a) Response time VMs allocated as load increases and released as it decreases a) Violation Rate Peaks occur a new VM is added Seminari DISI 2012 / Cloud Computing 55

56 Conclusions We described some recent results in the area of Cloud Computing Bologna View of Cloud Computing Energy reduction through VM consolidation P2P Clouds QoS management in Clouds Work in progress QoS management using performance runtime (e.g., QoS-aware energy reduction)... Cloud-enabled applications (mobility) Seminari DISI 2012 / Cloud Computing 56

57 Seminari DISI 2012 / Cloud Computing 57