Fabio Checconi

High Throughput Disk Scheduling with Fair Bandwidth Distribution

Mainstream applications-such as file copy/transfer, Web, DBMS, or video streaming-typically issue synchronous disk requests. As shown in this paper, this fact may cause work-conserving schedulers to fail both to enforce guarantees and provide a high disk throughput. A high throughput can be, however, recovered by just idling the disk for a short time interval after the completion of each request. In contrast, guarantees may still be violated by existing time-stamp-based schedulers because of the rules they use to tag requests. Budget Fair Queuing (BFQ), the new disk scheduler presented in this paper, is an example of how disk idling, combined with proper back-shifting of request time stamps, may allow a time-stamp-based disk scheduler to preserve both guarantees and a high throughput. Under BFQ, each application is always guaranteed-over any time interval and independently of whether it issues synchronous requests-a bounded lag with respect to its reserved fraction of the total number of bytes transferred by the disk device. We show the single-disk performance of our implementation of BFQ in the Linux kernel through experiments with real and emulated mainstream applications.

Providing performance guarantees to virtual machines using real-time scheduling

In this paper we tackle the problem of providing Quality of Service guarantees to virtualized applications, focusing on computing and networking guarantees. We propose a mechanism for providing temporal isolation based on a CPU real time scheduling strategy. This allows not only to have control over the individual virtual machine throughput, but also on the activation latency and response-time by which virtualized software components react to external events. We show experimental results gathered on a real system validating the approach.

Modular software architecture for flexible reservation mechanisms on heterogeneous resources

Management, allocation and scheduling of heterogeneous resources for complex distributed real-time applications is a challenging problem. Timing constraints of applications may be fulfilled by the proper use of real-time scheduling policies, admission control and enforcement of timing constraints. However, it is not easy to design basic infrastructure services that allow for easy access to the allocation of multiple heterogeneous resources in a distributed environment. In this paper, we present a middleware for providing distributed soft real-time applications with a uniform API for reserving heterogeneous resources with real-time scheduling capabilities in a distributed environment. The architecture relies on standard POSIX OS facilities, such as time management and standard TCP/IP networking services, and it is designed around CORBA, in order to facilitate modularity, flexibility and portability of the applications using it. However, real-time scheduling is supported by proper extensions at the kernel-level, plugged within the framework by means of dedicated resource managers. Our current implementation on Linux supports the reservation of the CPU, disk and network bandwidth. However, additional resource managers supporting alternative real-time schedulers for these resources, as well as additional types of resources, may be easily added. We present experimental results gathered on both synthetic applications and a real multimedia video streaming case study, showing the advantages deriving from the use of the proposed middleware. Finally, overhead figures are reported, showing the sustainability of the approach for a wide class of complex, distributed, soft real-time applications.

Real-time issues in live migration of virtual machines

This paper addresses the issue of how to meet the strict timing constraints of (soft) real-time virtualized applications while the Virtual Machine (VM) hosting them is undergoing a live migration. To this purpose, it is essential that the resource requirements of a migration are identified in advance, that appropriate resources are reserved to the process, and that multiple VMs sharing the same resources are temporally isolated from each other. The first issue is dealt with by introducing a stochastic model for the migration process. The other ones by introducing a methodology making use of proper scheduling algorithms (for both CPU and network) that allow for reserving resource shares to individual VMs. Also, an extensive set of simulations have been done by using traces of a VLC video server virtualized by using KVM on Linux. The traces have been obtained by patching KVM at the kernel level, and the same patch constitutes an important step towards the complete implementation of the proposed technique. The obtained results highlight the benefits of the proposed approach.

Self-tuning schedulers for legacy real-time applications

We present an approach for adaptive scheduling of soft real-time legacy applications (for which no timing information is exposed to the system). Our strategy is based on the combination of two techniques: 1) a real-time monitor that observes the sequence of events generated by the application to infer its activation period, 2) a feedback mechanism that adapts the scheduling parameters to ensure a timely execution of the application. By a thorough experimental evaluation of an implementation of our approach, we show its performance and its efficiency.

Virtualised e-Learning with real-time guarantees on the IRMOS platform

In this paper we focus on how Quality of Service guarantees are provided to virtualised applications in the Cloud Computing infrastructure that is being developed in the context of the IRMOS1 European Project. Provisioning of proper timeliness guarantees to distributed real-time applications involves the careful use of real-time scheduling mechanisms at the virtual-machine hypervisor level, of QoS-aware networking protocols and of proper design methodologies and tools for stochastic modelling of the application. The paper focuses on how we applied these techniques to a case-study involving a real e-Learning mobile content delivery application that has been integrated into the IRMOS platform and its achieved performance.

QFQ: efficient packet scheduling with tight guarantees

Packet scheduling, together with classification, is one of the most expensive processing steps in systems providing tight bandwidth and delay guarantees at high packet rates. Schedulers with near-optimal service guarantees and O(1) time complexity have been proposed in the past, using techniques such as timestamp rounding and flow grouping to keep their execution time small. However, even the two best proposals in this family have a per-packet cost component that is linear either in the number of groups or in the length of the packet being transmitted. Furthermore, no studies are available on the actual execution time of these algorithms. In this paper we make two contributions. First, we present Quick Fair Queueing (QFQ), a new O( 1) scheduler that provides near-optimal guarantees and is the first to achieve that goal with a truly constant cost also with respect to the number of groups and the packet length. The QFQ algorithm has no loops and uses very simple instructions and data structures that contribute to its speed of operation. Second, we have developed production-quality implementations of QFQ and of its closest competitors, which we use to present a detailed comparative performance analysis of the various algorithms. Experiments show that QFQ fulfills our expectations, outperforming the other algorithms in the same class. In absolute terms, even on a low-end workstation, QFQ takes about 110 ns for an enqueue()/dequeue() pair (only twice the time of DRR, but with much better service guarantees).

Virtualised e-Learning on the IRMOS real-time Cloud

This paper presents the real-time virtualised Cloud infrastructure that was developed in the context of the IRMOS European Project. The paper shows how different concepts, such as real-time scheduling, QoS-aware network protocols, and methodologies for stochastic modelling and run-time provisioning were practically combined to provide strong performance guarantees to soft real-time interactive applications in a virtualised environment. The efficiency of the IRMOS Cloud is demonstrated by two real interactive e-Learning applications, an e-Learning mobile content delivery application and a Virtual World e-Learning application.

The wizard of OS: a heartbeat for Legacy multimedia applications

Multimedia applications are often characterised by implicit temporal constraints but, in many cases, they are not programmed using any specialised real-time API. These “Legacy applications” have no way to communicate their temporal constraints to the OS kernel, and their quality of service (QoS), being necessarily linked to the temporal behaviour, fails to satisfy acceptable standards. In this paper we propose an innovative way for dealing with these applications, based on the combination of an on-line identification mechanism (which extracts from high-level observations such important parameters as the execution rate) and an adaptive scheduler (specialised for legacy applications) that identifies the correct amount of CPU needed by each application. Preliminary experimental results are reported, proving the effectiveness of the proposed idea in providing a widely used multimedia player on Linux with appropriate QoS guarantees, through an appropriate choice of the scheduling parameters. Finally, a detailed road-map is presented with the possible extensions to the approach.

W-CBS: a scheduling algorithm for supporting QoS in IEEE 802.11e

This paper presents a new scheduling algorithm, the Wireless Constant Bandwidth Server (W-CBS) for the Access Points of an IEEE 802.11e wireless networks to support traffic streams with Quality of Service guarantees, in particular in the case of multimedia applications which present variable bit rate traffic. The performance of W-CBS is compared to that of the reference scheduler defined in 802.11e standard using the ns2 simulator. The results show that the W-CBS outperforms the reference scheduler with VBR traffic, in terms of resource utilization and maximum admitted flows.