Presentation is loading. Please wait.

Presentation is loading. Please wait.

Progress Report Design, implementation, experiments, and demo plan 2014/12/03 1.

Similar presentations


Presentation on theme: "Progress Report Design, implementation, experiments, and demo plan 2014/12/03 1."— Presentation transcript:

1 Progress Report Design, implementation, experiments, and demo plan 2014/12/03 1

2 Project Goal Study the current scheduling algorithms for homogenous and asymmetric multi-core architecture. Design and implement the hypervisor scheduler on asymmetric multi-core platform. Assign virtual cores to physical cores for execution. Minimize the power consumption with performance guarantee. 2

3 3 Dom0 VCPU Xen Hypervisor Xen Hypervisor Asymmetry-aware Scheduler System Architecture Juno board Cortex- A53 Cortex- A57 Cortex- A53 Cortex- A57 Cortex- A53 DomU VCPU DomU VCPU

4 Main Work Items Functions that fetch physical and virtual core frequencies and send them to the scheduler. Implement the asymmetric-aware scheduler on Xen.

5 5 Dom0 VCPU Xen Hypervisor Xen Hypervisor Asymmetry-aware Scheduler Implementation - Setting DomU VCPU DomU VCPU Cortex- A53 Cortex- A57 Cortex- A53 Cortex- A57 Cortex- A53 One A53 core (called “master core”) is dedicated to Dom0.

6 6 Dom0 VCPU Xen Hypervisor Xen Hypervisor Asymmetry-aware Scheduler Implementation - Trigger Scheduling DomU VCPU DomU VCPU Cortex- A53 Cortex- A57 Cortex- A53 Cortex- A57 Cortex- A53 A timer is set on the master core during the initialization of the scheduler. It triggers the scheduler to generate plan every interval. Timer

7 7 Dom0 VCPU Xen Hypervisor Xen Hypervisor Asymmetry-aware Scheduler Implementation - Fetch frequency DomU VCPU DomU VCPU Cortex- A53 Cortex- A57 Cortex- A53 Cortex- A57 Cortex- A53 The scheduler fetches the frequency of each pcpu and vcpu before generating a scheduling plan.

8 8 Dom0 VCPU Xen Hypervisor Xen Hypervisor Asymmetry-aware Scheduler Implementation – Scheduler(Cont.) DomU VCPU DomU VCPU Cortex- A53 Cortex- A57 Cortex- A53 Cortex- A57 Cortex- A53 The scheduler generates a plan according to the frequency input, and assigns vcpus to the run queue of each pcpu.

9 Scheduler Implementation Build the scheduler based on the Xen scheduling framework. sched_asym.c

10 Some Technical Details Every physical core has a run queue. ◦ Contains a list of virtual cores, sorted by the start time. Every virtual core keeps a list of assignment. ◦ An assignment contains the information of target core, start and end time on target core. The scheduler updates the assignment of each vcpu according to the scheduling plan every time interval.

11 Example PCPU 1 PCPU 2 Vcpu 0 Vcpu 2 Vcpu 1 Vcpu 3 Target core:1 Start_time: 20 End_time :30 Target core: 2 Start_time: 50 End_time :60 Target core: 2 Start_time: 0 End_time :30 Target core: 1 Start_time: 0 End_time :20 … … Time = 0

12 Example PCPU 1 PCPU 2 Vcpu 0 Vcpu 2 Vcpu 1 Vcpu 3 Target core:1 Start_time: 20 End_time :30 Target core: 2 Start_time: 50 End_time :60 Target core: 2 Start_time: 0 End_time :30 Target core: 1 Start_time: 0 End_time :20 … … Time = 20 Vcpu 1 check its next plan_element

13 Example PCPU 1 PCPU 2 Vcpu 0 Vcpu 2 Vcpu 1 Vcpu 3 Target core:1 Start_time: 20 End_time :30 Target core: 2 Start_time: 50 End_time :60 Target core: 2 Start_time: 0 End_time :30 Target core: 2 Start_time: 30 End_time :50 … … Time = 20 Migrate to target core, and sort the run queue according to start time.

14 Example PCPU 1 PCPU 2 Vcpu 2 Vcpu 0 Vcpu 1 Vcpu 3 Target core:2 Start_time: 30 End_time :50 Target core: 2 Start_time: 50 End_time :60 Target core: 2 Start_time: 0 End_time :30 Target core: 1 Start_time: 20 End_time :30 … … Time = 20 Check if the next vcpu in the queue can start

15 Experiment Setting Juno board ◦ 1 Dom0: 1 virtual core  Xen ◦ 2 DomU: 1~4 virtual core(s)  VM(s) running benchmarks/applications. Probe ◦ Hardware: Juno Probe ◦ Software: Linaro Power Probe

16 Execution Scenario Same as the simulation scenarios in GCCE 2014 paper. ◦ Case 1: two virtual machines with light workload. ◦ Case 2: one virtual machine with heavy workload, the other with modestly workload. Benchmark ◦ CoreBench

17


Download ppt "Progress Report Design, implementation, experiments, and demo plan 2014/12/03 1."

Similar presentations


Ads by Google