Efficiently Scheduling Multi-Core Guest Virtual Machines on Multi-Core Hosts in Network Simulation

Abstract

Virtual machine (VM)-based simulation is a method used by network simulators to incorporate realistic application behaviors by executing actual VMs as high-fidelity surrogates for simulated end-hosts. A critical requirement in such a method is the simulation timeordered scheduling and execution of the VMs. Prior approaches such as time dilation are less efficient due to the high degree of multiplexing possible when multiple multi-core VMs are simulated on multi-core host systems. We present a new simulation time-ordered scheduler to efficiently schedule multi-core VMs on multi-core real hosts, with a virtual clock realized on each virtual core. The distinguishing features of our approach are: (1) customizable granularity of the VM scheduling time unit on the simulation time axis, (2) ability to take arbitrary leaps in virtual time by VMs to maximize the utilization of host (real) cores when guest virtual cores idle, and (3) empirically determinable optimality in the tradeoff between total execution (real) time and time-ordering accuracy levels. Experiments show that it is possible to get nearly perfect time-ordered execution, with a slight cost in total run time, relative to optimized non-simulation VM schedulers.Interestingly, with our time-ordered scheduler, it is also possible to reduce the time-ordering error from over 50% of non-simulation scheduler to less than 1% realized by our scheduler, with almost the same run time efficiency as that of the highly efficient nonsimulation VM schedulers.

Dynamic Deadline Aware Efficient Virtual Machine Scheduling in the Cloud

“…Zhiyun Zheng [7] proposed new algorithm based on shered shared waiting queue. Yoginath [6] proposed new scheduling algorithm based on time-order. Chen [1] Designed Affinity-aware Proportional Share Scheduling for Virtual Machine.…”

Design of a high-fidelity testing framework for secure electric grid control

“…In the rest of the paper, the terms telecommunication network simulator and network simulator are synonymously used. High-fidelity time-accurate network simulation systems such as NetWarp (Yoginath and Perumalla, 2011;Yoginath, Perumalla, and Henz 2012), based on VMs are extremely attractive for the task of including networked VMs to model grid control networks. However, such network simulation test-beds for distributed protocols and applications are not directly amenable to work in tandem with a electric grid simulator.…”

“…In order to ensure correctness of the simulation time-ordered scheduling of VMs becomes necessary (Yoginath and Perumalla 2011;Yoginath, Perumalla, and Henz 2012). To overcome this, the virtual time for each VM based on the number of CPU cycles utilized is recorded by each VCPU of the VM and this value is the LVT in the network simulator context.…”

Optimized Hypervisor Scheduler for Parallel Discrete Event Simulations on Virtual Machine Platforms

“…Next, we need to implement the hypervisor scheduler that employs a least-simulation-time-first policy instead of the default credit-based fair scheduling strategy. Implementing the Xen hypervisor scheduler for the application-specific requirements has been previously accomplished [7] [8]. Each PCPU maintains a runq (priority-queue) in which the VCPUs requiring clock-cycles are en-queued.…”

“…The PDES Scheduler for Xen (PSX) scheduler replaces the default Credit Scheduler of Xen (CSX) in scheduling the virtual CPU (VCPUs) onto the physical cores of CPU (PCPU). The strategy that we take to replace the scheduler is similar to the one presented by [8].…”