Pavan Kumar Paluri
Albert Mo Kim Cheng
ABSTRACT
A multitude of cloud-native applications take up a significant share of today's world wide web, the majority of which implicitly require soft-real-time guarantees when hosted on servers at various data centers across the globe. With the rapid development of cloud computing and virtualization techniques, many applications have been moved onto cloud and edge platforms that require efficient virtualization techniques. This means a set of applications must be executed on a Virtual Machine (VM) and multiple VMs must be temporally and spatially scheduled on a set of CPUs. Designed to leverage the cloud infrastructure model, many of these cloud-native applications such as media servers strongly demand low data latency and high compute-resource availability, both of which must be predictable. However, state-of-art VM schedulers fail to satisfy these requirements simultaneously. The scheduling of cloud-native applications on VMs and the scheduling of VMs on physical resources (CPUs), collectively need to be real-time in nature as specified by the Hierarchical Real-Time Scheduling (HiRTS) framework. Conforming to the specifications of this framework, the Regularity-based Resource Partitioning (RRP) model has been proposed that introduces the concept of regularity to provide a near-ideal resource supply to all VMs. In this paper, we make the theoretically superior Regularity-based Resource Partitioning (RRP) model ready for prime time by implementing its associated resource partitioning algorithms for the first time ever on the popular x-86 open-source hypervisor Xen, i.e., RRP-Xen. This paper also compares and contrasts the real-time performance of RRP-Xen against contemporary Xen schedulers such as Credit and RTDS. Our contributions include: (1) a novel implementation of the RRP model on Xen's x-86 based hypervisor, thereby providing a test-bed for future researchers; (2) the first-ever multi-core ARINC 653 VM scheduler prototype on Xen; and (3) numerous experiments and theoretical analysis to determine the real-time performance of RRP-Xen under a stringent workload environment.
- [n.d.]. wrk2: A constant throughput, correct latency recording variant of wrk.. EB/OL. https://doi.org/10.5281/zenodo.4796378 Google Scholar
Cross Ref
- Paul Barham, Boris Dragovic, Keir Fraser, Steven Hand, Tim Harris, Alex Ho, Rolf Neugebauer, Ian Pratt, and Andrew Warfield. 2003. Xen and the art of virtualization. In ACM SIGOPS operating systems review. 37, 164–177. https://doi.org/10.1145/1165389.945462 Google ScholarDigital Library
- Allan Borodin, Denis Pankratov, and Amirali Salehi-Abari. 2018. A Simple PTAS for the Dual Bin Packing Problem and Advice Complexity of Its Online Version. In 1st Symposium on Simplicity in Algorithms (SOSA 2018), Raimund Seidel (Ed.) (OpenAccess Series in Informatics (OASIcs), Vol. 61). Schloss Dagstuhl–Leibniz-Zentrum fuer Informatik, Dagstuhl, Germany. 8:1–8:12. isbn:978-3-95977-064-4 issn:2190-6807 https://doi.org/10.4230/OASIcs.SOSA.2018.8 Google ScholarCross Ref
- John M. Calandrino, Hennadiy Leontyev, Aaron Block, UmaMaheswari C. Devi, and James H. Anderson. 2006. LITMUS\^ RT : A Testbed for Empirically Comparing Real-Time Multiprocessor Schedulers. In 2006 27th IEEE International Real-Time Systems Symposium (RTSS’06). 111–126. https://doi.org/10.1109/RTSS.2006.27 Google ScholarDigital Library
- Wei-Ju Chen, Peng Wu, Pei-Chi Huang, Aloysius K Mok, and Song Han. 2019. Online Reconfiguration of Regularity-Based Resource Partitions in Cyber-Physical Systems. In 2019 IEEE Real-Time Systems Symposium (RTSS). 495–507. https://doi.org/10.1007/s11241-021-09364-5 Google ScholarCross Ref
- Albert M. K. Cheng, Guangli Dai, Pavan Kumar Paluri, Mansoor Ansari, Darrel Knape, and Yu Li. 2019. Fault-Tolerant Regularity-Based Real-Time Virtual Resources. In 2019 IEEE 25th International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA). 1–12. https://doi.org/10.1109/RTCSA.2019.8864575 Google ScholarCross Ref
- Hyunsik Choi, Saeyoung Han, Sungyong Park, and Eunji Yang. 2009. A CPU provision scheme considering virtual machine scheduling delays in Xen virtualized environment. In TENCON 2009 - 2009 IEEE Region 10 Conference. 1–6. https://doi.org/10.1109/TENCON.2009.5396059 Google ScholarCross Ref
- Guangli Dai, Pavan Kumar Paluri, and Albert M. K. Cheng. 2018. RRP Edge Computing System. In International Workshop on Trustworthy & Real-time Edge Computing for Cyber-Physical Systems. TN, USA. https://doi.org/10.5281/zenodo.4794272 Google ScholarCross Ref
- Guangli Dai, Pavan Kumar Paluri, and Albert M. K. Cheng. 2018. Task Mapping in a Regularity-based Resource Partitioning Hierarchical Real-Time System. In Work-in-Progress Proceedings of 30th Euromicro Conference on Real-Time Systems. Barcelona, Spain. https://doi.org/10.5281/zenodo.4794478 Google ScholarCross Ref
- Arvind Easwaran, Madhukar Anand, and Insup Lee. 2007. Compositional Analysis Framework Using EDP Resource Models. In Proceedings of the 28th IEEE International Real-Time Systems Symposium. USA. 129–138. isbn:0769530621 https://doi.org/10.1109/RTSS.2007.17 Google ScholarDigital Library
- Yu Li and Albert M. K. Cheng. 2012. Static Approximation Algorithms for Regularity-based Resource Partitioning. In 2012 IEEE 33rd Real-Time Systems Symposium. 137–148. https://doi.org/10.1109/RTSS.2012.66 Google ScholarDigital Library
- Yu Li and Albert Mo Kim Cheng. 2016. Transparent Real-Time Task Scheduling on Temporal Resource Partitions. IEEE Trans. Comput., 65, 5 (2016), 1646–1655. https://doi.org/10.1109/TC.2015.2449857 Google ScholarDigital Library
- Yu Li and Albert M. K. Cheng. 2017. Toward a Practical Regularity-Based Model: The Impact of Evenly Distributed Temporal Resource Partitions. ACM Trans. Embed. Comput. Syst., 16, 4 (2017), Article 111, Aug., 25 pages. issn:1539-9087 https://doi.org/10.1145/3092945 Google ScholarDigital Library
- Tiago Macedo and Fred Oliveira. 2011. Redis Cookbook: Practical Techniques for Fast Data Manipulation. " O’Reilly Media, Inc.". isbn:9781449305048 isbn:9781449305048.Google Scholar
- A.K. Mok and Xiang Alex. 2001. Towards compositionality in real-time resource partitioning based on regularity bounds. In Proceedings 22nd IEEE Real-Time Systems Symposium (RTSS 2001) (Cat. No.01PR1420). 129–138. https://doi.org/10.1109/REAL.2001.990604 Google ScholarCross Ref
- Will Reese. 2008. Nginx: the high-performance web server and reverse proxy. Linux Journal, 2008, 173 (2008), 2. https://web.archive.org/web/20210116123712/https://www.linuxjournal.com/article/10108Google ScholarDigital Library
- Insik Shin and Insup Lee. 2003. Periodic resource model for compositional real-time guarantees. In RTSS 2003. 24th IEEE Real-Time Systems Symposium, 2003. 2–13. https://doi.org/10.1109/REAL.2003.1253249 Google ScholarCross Ref
- Rahul Soni. 2016. Nginx. Springer. https://doi.org/10.1007/978-1-4842-1656-9 Google ScholarCross Ref
- Salvador Trujillo, Alfons Crespo, and Alejandro Alonso. 2013. MultiPARTES: Multicore Virtualization for Mixed-Criticality Systems. In 2013 Euromicro Conference on Digital System Design. 260–265. https://doi.org/10.1109/DSD.2013.37 Google ScholarDigital Library
- Steven H. VanderLeest. 2010. ARINC 653 hypervisor. In 29th Digital Avionics Systems Conference. 5.E.2–1–5.E.2–20. https://doi.org/10.1109/DASC.2010.5655298 Google ScholarCross Ref
- Manohar Vanga, Arpan Gujarati, and Björn B. Brandenburg. 2018. Tableau: A High-Throughput and Predictable VM Scheduler for High-Density Workloads. In Proceedings of the Thirteenth EuroSys Conference (EuroSys ’18). Association for Computing Machinery, New York, NY, USA. Article 28, 16 pages. isbn:9781450355841 https://doi.org/10.1145/3190508.3190557 Google ScholarDigital Library
- Meng Xu, Linh T.X. Phan, Insup Lee, Oleg Sokolsky, Sisu Xi, Chenyang Lu, and Christopher Gill. 2013. Cache-Aware Compositional Analysis of Real-Time Multicore Virtualization Platforms. In 2013 IEEE 34th Real-Time Systems Symposium. 1–10. https://doi.org/10.1109/RTSS.2013.9 Google ScholarDigital Library
Index Terms
- ARINC 653-inspired regularity-based resource partitioning on xen
Recommendations
RT-Xen: towards real-time hypervisor scheduling in xen
EMSOFT '11: Proceedings of the ninth ACM international conference on Embedded softwareAs system integration becomes an increasingly important challenge for complex real-time systems, there has been a significant demand for supporting real-time systems in virtualized environments. This paper presents RT-Xen, the first real-time hypervisor ...
Virtio network paravirtualization driver
One of the techniques used to improve I/O performance of virtual machines is paravirtualization. Paravirtualized devices are intended to reduce the performance overhead on full virtualization where all hardware devices are emulated. The interface of a ...
Improving Agility and Elasticity in Bare-metal Clouds
ASPLOS '15: Proceedings of the Twentieth International Conference on Architectural Support for Programming Languages and Operating SystemsBare-metal clouds are an emerging infrastructure-as-a-service (IaaS) that leases physical machines (bare-metal instances) rather than virtual machines, allowing resource-intensive applications to have exclusive access to physical hardware. Unfortunately,...
Comments