Evaluating industrial applicability of virtualization on a distributed multicore platform

Abstract

Adoption of virtualization technology has been limited in industrial automation due to unavailability of mature solutions, and strict timing requirements of control systems. However, current advancement in Virtual Monitoring Machine, multicore technology, virtualization extension and network virtualization has led to increased interest of virtualization in industrial automation.So far, many related research are focused on maximizing CPU and I/O utilization, and optimization applicable to soft realtime systems (i.e., outside industrial automation domain), e.g., multimedia applications. In this research, we make use of QoS for CPU, memory and network bandwidth in pursuit of high speed and predictability on a distributed multicore platform which is constructed entirely from open source products. We evaluate the platform for latency and jitter, network throughput and CPU computation load. Finally, we analyze the result for applicability in industrial control domain.

State Management for Cloud-Native Applications

“…In contrast, our proposed state layer could operate in a single-cloud infrastructure handling a large number of small pieces of data (internal states/variables, such as integers and strings) which are accessed according to a constantly changing and dynamic NF reads/writes characteristic. In addition, the accesses should take no longer than a few milliseconds [13].…”

“…However, despite the cloud-native concept's promises, eliminating the need for explicit state management within applications does not come for free; the downside is the complexity faced by all distributed systems, the performance, and data consistency [1][2][3][4]. These difficulties are especially compelling in telco deployments [11,12], where a carrier-grade telco NF ecosystem must provide extreme steady-state performance and strict service-level objectives (SLOs) simultaneously, at the order of hundreds of millions of packets per second throughput, and a 1-10 ms delay budget [13]. Although the perceived performance overhead of state externalization depends on various factors (such as the application's state access pattern, the performance of the backing data store, and the deployment characteristics), we can anticipate that the cost of external state access may easily become the dominating, that is, limiting, factor in the overall NF performance.…”

Towards Mission-Critical Control at the Edge and Over 5G

“…Other works attempt to characterise and profile the performance of different aspects of the edge cloud. For example, the authors of [8] evaluate a generic platform for industrial control, with respect to latency, throughput, and CPU load. Their focus is on the pros and cons of virtualisation in a 'Multi-core' environment rather than the 'cloud'.…”

Control of robots and machine tools with an extended factory cloud

“…Mahmud, Sandstrom and Vulgarakis [8] approach this problem by building a distributed multicore platform from open-source products and evaluating it for latency, jitter, network throughput and CPU computation load. There are efforts by project pICASSO [9] and Verl et al [10] to build a full-scale industrial cloud platform for cyber-physical production systems (CPPS).…”