Real-Time Cloud Computing Chenyang Lu Cyber-Physical Systems Laboratory http://www.cse.wustl.edu/~lu/
Internet of Things Convergence of Ø Miniaturized hardware: integrate processor, sensors and radios. Ø Low-power wireless: connect millions of devices to the Internet. Ø Data analytics: make sense of sensor data. Ø Cloud: scalable computing. R. Dor, G. Hackmann, Z. Yang, C. Lu, Y. Chen, M. Kollef and T.C. Bailey, Experiences with an End-To- End Wireless Clinical Monitoring System, Conference on Wireless Health (WH'12), October 2012.
Internet of Things Convergence of Ø Miniaturized hardware: integrate processor, sensors and radios. Ø Low-power wireless: connect millions of devices to the Internet. Ø Data analytics: make sense of sensor data. Ø Cloud: scalable computing. R. Dor, G. Hackmann, Z. Yang, C. Lu, Y. Chen, M. Kollef and T.C. Bailey, Experiences with an End-To- End Wireless Clinical Monitoring System, Conference on Wireless Health (WH'12), October 2012.
Real-Time Cloud for CPS Ø Large-scale IoT -driven control q Smart manufacturing, transportation, infrastructure… q Closed-loop control à real-time performance q Computing at scale à cloud Ø Real-time cloud: enabling technology for large CPS!
Smart City Ø Manage assets intelligently through large-scale IoT -driven control Ø Example: Intelligent Transportation q Collect data from roadside sensors and cameras q Centralized data analysis q Control city-wise traffic signals intelligently q SCATS @ Sydney [1]: controlling 3,400 signals at 1s latency Ø Many concurrent connections Ø Require low latency communication [1] http://en.wikipedia.org/wiki/Traffic_light_control_and_coordination 11/7/19 5
Smart Civil Infrastructure Cyber-Physical Boundary WU/Purdue project in Real-Time Hybrid Simulation - Enabled by real-time parallel computing - Expand to larger-scale, multi-specimen experiments (bridge spanning a river, different ground motions on each end) - Towards cloud-based multi-site experiments 6
Industrial Internet of Things (IIoT) Ø Differentiated real-time and reliability requirements q Latency q Delivery guarantees q Event time consistency at-least-once best-effort milliseconds e.g., Emergency response e.g., Real-time monitoring hours e.g., Predictive maintenance 11/7/19 7
Embedded System Virtualization Ø Consolidate 100 ECUs à ~10 multicore processors. Ø Integrate multiple systems on a common platform. q Infotainment on Linux or Android q Safety-critical control on AUTOSAR Ø Preserve real-time performance on a virtualized platform! Source: http://www.edn.com/design/automotive/4399434/Multicore-and-virtualization-in-automotive-environments 11/7/19 8
Cloud is real-time today Ø Virtualization platforms provide no guarantee on latency q Xen: credit scheduler, [credit, cap] q VMware ESXi: [reservation, share, limitation] q Microsoft Hyper-V: [reserve, weight, limit] Ø Clouds lack service level agreement on latency q Amazon, Google, Microsoft cloud services: #VCPUs Current clouds provision resources, not latency! 9
Real-Time Cloud Ø Support real-time applications in the cloud. q Latency guarantees for tasks running in virtual machines (VMs). q Real-time performance isolation between VMs. q Resource sharing between real-time and non-real-time VMs. Ø Real-time cloud stack. q RT -Xen à real-time VM scheduling q VATC à real-time network I/O on a virtualized host. q RT -OpenStack à real-time cloud resource management. Cyber-Physical RT Cilk Plus Event Processing -OpenStack VATC: RT Network I/O Latency guarantees RT 10
Xen Ø Xen: type-1, baremetal hypervisor q Domain-0: drivers, tool stack to control VMs. q Guest Domain: para-virtualized or fully virtualized OS. Ø Scheduling hierarchy q Xen schedules VCPUs on PCPUs. q Guest OS schedules threads on VCPUs. q Xen credit scheduler: round-robin with proportional share. VCPU Real-Time Task OS Sched OS Sched OS Sched Xen Scheduler PCPUs 11
RT-Xen Ø Real-time schedulers in the Xen hypervisor. Ø Provide real-time guarantees to tasks in VMs. Ø Incorporated in Xen 4.5 as the real-time scheduler. RT-Xen https://sites.google.com/site/realtimexen/ S. Xi, M. Xu, C. Lu, L. Phan, C. Gill, O. Sokolsky and I. Lee, Real-Time Multi-Core Virtual Machine Scheduling in Xen, ACM International Conference on Embedded Software (EMSOFT'14), October 2014. 12
Compositional Scheduling Ø Analytical real-time guarantees to tasks running in VMs. Ø VM resource interfaces q A set of VCPUs each with resource demand <period, budget > q Hides task-specific information q Computed based on compositional scheduling analysis Hypervisor Resource Interface Scheduler Resource Interface Resource Interface Scheduler Scheduler Workload Workload Virtual Machines 13
Real-Time Scheduler Design Ø Global scheduling q Allow VCPU migration across cores q Work conserving – utilize any available cores q Migration overhead and cache penalty Ø Partitioned scheduling q Assign and bind VCPUs to cores q Cores may idle when others have work pending q No migration overhead or cache penalty Ø Enforce resource interface through budget management q Periodic server vs. deferrable server Ø Priority: Earliest Deadline First vs. Deadline Monotonic 14
VCPU Scheduled as a Deferrable Server (0,4) (0,4) (0,4) (7,4) (13,4) ... Task Schedule Task Schedule Task Schedule Task Schedule t t t t 0 0 0 0 4 4 4 4 8 8 8 8 12 12 12 12 16 16 16 16 3 3 3 3 VCPU Schedule VCPU Schedule VCPU Schedule VCPU Schedule Budget Budget Budget Budget ... Budget = 3 Budget = 3 Budget = 3 Budget = 3 Period = 4 Period = 4 Period = 4 Period = 4 t t t t 0 4 8 12 16 16 16 16 0 0 0 4 4 4 8 8 8 12 12 12 Ø A Deferrable Server has two parameters < budget, period> q The server consumes budget when executing jobs q When the budget exhausted, the server stops executing jobs q Budget replenishes at the start of each period 15 [1] Xi, Sisu, et al. "Real-time multi-core virtual machine scheduling in xen." 2014 International Conference on Embedded Software (EMSOFT) . IEEE, 2014. [2] Kim, Hyoseung, Shige Wang, and Ragunathan Rajkumar. "vMPCP: A synchronization framework for multi-core virtual machines." 2014 IEEE Real-Time Systems Symposium . IEEE, 2014
RT-Xen vs. Xen • Xen misses deadlines at 22% of CPU capacity. • RT -Xen delivers real-time performance at 78% of CPU capacity. 16
Virtualized Network I/O Ø Xen handles all network traffic through Dom0 Ø Real-time and non-real-time traffic share Dom0 q CPU and network contention Ø Long delays for real-time traffic in virtualized hosts Dom0 Dom1 Dom2 Network Non Real-Time Real-Time Components App App Xen Hypervisor CPU Memory Storage NIC 17
Network I/O in Virtualized Hosts Dom1 Dom2 Ø Linux Queueing Discipline Non- Real- q Rate-limit and shape flows Real- Time Time App App q Prioritization or fair packet scheduling Ø Priority inversion in virtualization Virtualization Components components Dom0 q between transmissions q between transmission and reception Queueing Discipline Ø VATC: Virtualization-Aware Traffic Control q Process packets in prioritized kernel threads q Dedicated packet queues per priority NIC C. Li, S. Xi, C. Lu, C. Gill and R. Guerin, Prioritizing Soft Real-Time Network Traffic in Virtualized Hosts Based on Xen, IEEE Real-Time and Embedded Technology and Applications Symposium (RTAS'15), April 2015. 18
Real-Time Traffic Latency VATC reduces priority inversion à lower latency for real-time traffic. 4 Prio, Dom0 − 3.18 3.5 FQ_CoDel, Dom0 − 3.18 VATC Round − trip Latency ( ms) 3 2.5 2 1.5 1 0.5 0 10 16 32 64 128 256 512 1024 Dyn Cons Dyn Interrupt Interval ( µ s) • Median round-trip latency of real-time traffic. • CPU contention from two small-packet interfering streams. 19
Virtualized Host à Cloud Ø Provide real-time performance to real-time VMs Ø Achieve high resource utilization 20
OpenStack Limitations Ø Popular open-source cloud management system Manager Ø VM resource interface q Number of VCPUs Host Host Host q Not real-time VM VM VM VM VM Ø VM-to-host mapping q Filtering (admission control) • VCPU-to-PCPU ratio (16:1), max VMs per host (50) • Coarse-grained admission control for CPU resources q Ranking (VM allocation) • Balance memory usage • No consideration of CPU resources 21
RT-OpenStack Ø Co-hosting real-time VMs with non-real-time VMs Ø Deliver real-time performance q Support RT -Xen resource interface q Real-time-aware VM-to-host mapping Ø Achieve high resource utilization q Co-locate non-real-time VMs with real-time VMs q Non-real-time VMs consume remaining resources without affecting the real-time performance of real-time VMs S. Xi, C. Li, C. Lu, C. Gill, M. Xu, L. Phan, I. Lee, O. Sokolsky, RT-OpenStack: CPU Resource Management for Real- Time Cloud Computing, IEEE International Conference on Cloud Computing (CLOUD'15), June 2015. 22
RT-OpenStack: VM-to-Host Mapping Ø Admission control: RT -Filter q Accept real-time VMs based on schedulability and memory q Consider only accepted real-time VMs Ø VM allocation: RT -Weigher q Balance CPU utilization q Consider only accepted real-time VMs Resource Interface Admission Control VM Allocation Schedulability + Real-Time VMs {<period, budget>} CPU Utilization Memory Non-Real-Time VMs Best Effort Memory Memory 23
Recommend
More recommend
