Cost-Efficient Resource Management for Scientific Workflows on the - PowerPoint PPT Presentation

Cost-Efficient Resource Management for Scientific Workflows on the Cloud Ilia Pietri School of Computer Science, University of Manchester, U.K

Overview Scientific Meet User User Workflows Constraints Virtual Machines Goals (VMs) Provide Hosts Cloud Provider Energy Efficiency 1/21

Problem Description • Scientific Workflows A – DAGs* – Data dependency constraints C D B – Gaps in the schedule • Goal: F E – Increase resource utilisation and achieve cost-efficient provisioning G • Estimation of resource needs • CPU frequency selection *DAG: Directed Acyclic Graph 2/21

Motivation • A wide range of possible configurations – Number of resources – CPU frequency • New pricing schemes – Pricing for CPU provisioning – Faster resources cost more • Which option to choose? – Extra resources at a lower frequency – Less resources but faster 3/21

Motivation 4/21

Motivation 5/21

Determining the Amount of Resources • Assuming the user is interested in executing a scientific workflow – How many resources to provision? – Cost vs performance User: cost-efficient execution, as quickly as possible Provider: minimize energy costs – Performance modelling estimate workflow execution time (and related costs) for a different number of slots 6/21

A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud • Level-based Estimation Model A – Workflow structure and individual job characteristics C B D – Task assignment to levels • Top-Down Approach • Bottom-Up Approach F E – Overall workflow performance estimation G • Based on level characteristics H Pietri I., Juve G., Deelman E., Sakellariou R., A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud. In 9 th WORKS @ SC14. 2014. 7/21

A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud • Level-based Estimation Model A – Workflow structure and individual job characteristics C B D – Task assignment to levels • Top-Down Approach • Bottom-Up Approach F E – Overall workflow performance estimation G • Based on level characteristics H Pietri I., Juve G., Deelman E., Sakellariou R., A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud. In 9 th WORKS @ SC14. 2014. 7/21

A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud • Level-based Estimation Model A – Workflow structure and individual job characteristics C B – Task assignment to levels • Top-Down Approach • Bottom-Up Approach E D – Overall workflow performance estimation G F • Based on level characteristics H Pietri I., Juve G., Deelman E., Sakellariou R., A Performance Model to Estimate Execution Time of Scientific Workflows on the Cloud. In 9 th WORKS @ SC14. 2014. 7/21

Performance Modelling for Montage 8/21

Using the model in practice • Estimation of makespan for – Different number of slots – Different CPU frequencies runtime f =[ β t *(f/f max -1)+1]*runtime fmax • Use of estimated makespan to predict – User monetary costs – Energy costs 9/21

Improving the Schedule for Energy Savings • The user is interested in executing a workflow within a deadline • Given an initial schedule* – Lower the CPU frequency to execute the tasks and achieve energy savings Utilising the slack time * an allocation of the tasks to the resources 10/21

Slack Time • Slack time : Maximum delay in task execution without exceeding the deadline * − slackTime t =min(spareTime t->s + slackTime s ), s ϵ succ t For the exit node: − slackTime texit =deadline-finishTime texit − Where spare time is the maximum delay in task execution without affecting the successors start times [1] spareTime t =startTime s -finishTime t -comCost t->s [1] R. Sakellariou and H. Zhao, “A low -cost rescheduling policy for efficient mapping of workflows on grid systems,” Scientific Programming , vol. 12, no. 4, pp. 253 – 262, 2004 11/21

Algorithms with DVFS Techniques Example P0 P1 P2 0 1 3 2 4 5 6 7 8 12/21

Algorithms with DVFS Techniques Example P0 P1 P2 0 1 3 2 4 5 6 Data transfer 6->8 finishes 7 Execution of task 8 finishes 8 Execution of task 8 starts Deadline 12/21

Algorithms with DVFS Techniques Example P0 P1 P2 0 1 3 2 4 5 6 Penalty on execution time 7 8 Start time of task 8 is affected 12/21

Energy Savings • Frequency scaling may not always be energy-efficient – Power savings, but longer execution time: runtime f =[ β t *(f/f max -1)+1]*runtime fmax – Idle power to be considered P fmax P idle time idle runtime fmax P f runtime f [1] C.H. Hsu, U. Kremer . The design, implementation, and evaluation of a compiler algorithm for cpu energy reduction. ACM SIGPLAN Notices , 38(5):38 – 48, 2003. [2] M . Etinski, Corbalan, J. Labarta, and M. Valero, “Understanding the future of energy-performance trade-off via DVFS in HPC 13/21 environments,” JPDC , vol. 72, no. 4, pp. 579 – 590, 2012.

Idea • Workflows may consist of heterogeneous tasks – Different energy vs frequency behaviour – Reducing the frequency of a task may impact overall schedule • Idea – Start with an initial schedule e.g. HEFT [1] – Apply frequency scaling iteratively • Based on the energy gain of each task • Assessing the impact on overall energy consumption [1] H. Topcuoglu, S. Hariri, and M.- Y. Wu, “Performance -effective and low-complexity task scheduling for heterogeneous computing,” IEEE TPDS , vol. 13, no. 3, pp. 260 – 274, 2002 14/21

Energy-Aware Workflow Scheduling Using Frequency Scaling ESFS (Energy-aware Stepwise Frequency Scaling) • Stage 1. Start with an initial schedule at maximum frequency. • Stage 2. Using the next available frequency (bound): − 1. Identify the most profitable tasks in terms of energy gain(beyond a threshold). − 2. Update the plan using the lower frequencies for these tasks. − 3. Assess the impact to overall energy consumption (for the whole workflow). − 4. Continue with the procedure as long as energy savings are achieved (stage 2). I. Pietri, R. Sakellariou . “Energy -Aware Workflow Scheduling Using Frequency Scaling ”. ICPP Workshops (PASA), 2014. 15/21

SIPHT with 1000 tasks [1] I. Pietri, R. Sakellariou . “Energy - Aware Workflow Scheduling Using Frequency Scaling”. ICPP Workshops (PASA), 2014. [2] Q. Huang et al. “Enhanced energy- efficient scheduling for parallel applications in cloud,” in Proceedings of the 12th IEEE/ACM CCGrid . IEEE, 2012, pp. 781 – 786. 16/21

Selecting CPU Frequencies Per Resource • We proposed an energy-efficient approach that selects the CPU frequency for each task. • In practice, we may need to choose a CPU frequency per resource (or core). – CPU provisioning charged based on frequency • Linear relation between frequency and price e.g. CloudSigma [1] and ElasticHosts [2] – User: minimize the cost within a deadline [1] https://www.cloudsigma.com/ [2] http://www.elastichosts.com/ 17/21

Cost-efficient Provisioning of Cloud Resources Priced by CPU CSFS (Cost-based Stepwise Frequency Selection) • Stage 1. Start with a schedule at maximum frequency. • Stage 2. Using the next available frequency: − 1. Select the resources with cost savings for the new frequency. − 2. Update the plan using the chosen frequency for these resources. − 3. Accept the new plan if costs savings and continue with the same procedure (go to 1). − 4. Otherwise terminate. I. Pietri and R. Sakellariou. Cost-Efficient Provisioning of Cloud Resources Priced by CPU frequency. Best poster award at UCC2014, 2014. 18/21

Montage with 1000 tasks 19/21

Pricing in practice • How to use pricing to motivate users for energy-efficient scheduling in practice? − Trade-off between • Energy savings for the provider • Minimization of user cost D. Lucanin, I. Pietri, I. Brandic and R. Sakellariou. A Cloud Controller for Performance-Based Pricing. In IEEE Cloud 2015 , to appear. 20/21

A Cloud Controller for Performance-based Pricing • Motivation – The impact on application performance from frequency reduction may vary depending on the application characteristics • Perceived-performance pricing – CPU provisioning charged based on impact on performance • CPU frequency used • VM CPU-boundedness • Two-stage cloud controller – Allocation of VMs to PMs – CPU frequency scaling when energy savings exceed the revenue losses D. Lucanin, I. Pietri, I. Brandic and R. Sakellariou. A Cloud Controller for Performance-based Pricing. In IEEE Cloud 2015 , to appear. 21/21

Cost-Efficient Resource Management for Scientific Workflows on the Cloud Thank you for your time!