A measurement-based analysis of the real-time performance of linux

Abeni, Luca; Goel, Ashvin; Krasic, Charles; Snow, Jim; Walpole, Jonathan

doi:10.1109/rttas.2002.1137388

Cited by 73 publications

(64 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Multimedia applications, and more generally soft realtime applications, are characterized by implicit temporal constraints that must be met to provide the desired QoS [17]. Assuming that tracing the kernel scheduler has a negligible impact on the system, we can verify that temporal constraints are satisfied for one or multiple real-time applications, and whenever they are not, we can show what the system was doing at that time.…”

Section: Performance Debuggingmentioning

confidence: 99%

“…However, this shows how our approach was useful to identify suspicious code sequences. To support soft real-time applications, the kernel should respect the application's temporal constraints and therefore a predictable schedule is desired [17]. Such applications may require periodic scheduling where the period is derived from the frame rate of an audio/video stream, for example.…”

Section: Locking Validationmentioning

confidence: 99%

See 1 more Smart Citation

Operating System Level Trace Analysis for Automated Problem Identification

Matni¹,

Dagenais²

2011

TOCSJ

View full text Add to dashboard Cite

Performance bottlenecks, malicious activities, programming bugs and other kinds of problematic behavior could be accurately detected on production systems if the relevant events were being monitored. This could be achieved through kernel level tracing where every time a relevant event occurs, the information is analysed or saved in a trace file to be inspected during post-mortem analysis. While collecting the information from the kernel has a very low impact, the offline analysis is typically performed remotely with no overhead on the system whatsoever.This article presents an automata-based approach for analyzing traces generated by the kernel of an operating system. Some typical patterns of problematic behavior are identified and described using the State Machine Language. These patterns are fed into an offline analyzer which efficiently and simultaneously checks for their occurrences even in traces of several gigabytes. The analyzer achieves a linear performance with respect to the trace size. The remaining factors impacting its performance are also discussed. The main interest of the proposed approach is the efficiency obtained in monitoring such extensive and detailed execution traces for a very large number of simultaneous possible patterns of problematic behavior.

show abstract

Section: Performance Debuggingmentioning

confidence: 99%

Section: Locking Validationmentioning

confidence: 99%

Operating System Level Trace Analysis for Automated Problem Identification

Matni¹,

Dagenais²

2011

TOCSJ

View full text Add to dashboard Cite

show abstract

“…The real-time patch is a modification to a plain Linux kernel to support kernel preemption [5]. It achieves a few hundred µseconds latency [6], but still the result is slower by a factor of ten comparing to typical RTOSes. Even though the mechanism is potentially capable of achieving real-time responsiveness, it could be easily spoiled by a bad-mannered device driver, which holds a lock for a long period.…”

Section: Related Workmentioning

confidence: 99%

Constructing a Multi-OS Platform with Minimal Engineering Cost

Kinebuchi

Morita

Makijima

et al. 2009

IFIP Advances in Information and Communication Technology

View full text Add to dashboard Cite

Abstract. Constructing an embedded device with a real-time and a general-purpose operating system has attracted attention as a promising approach to let the device balance real-time responsiveness and rich functionalities. This paper introduces our methodology for constructing such multi-OS platform with minimal engineering cost by assuming asymmetric OS combinations unique to embedded systems. Our methodology consists of two parts. One is a simple hypervisor for multiplexing resources to be shared between operating systems. The other is modifying operating systems to allow them to be aware of each other. We constructed an experimental system executing TOPPERS and Linux simultaneously on a hardware equipped with an SH-4A processor. The modification to each operating system kernel limited to a few dozen lines of code and do not introduce any overhead that would compromise real-time responsiveness or system throughput.

show abstract

“…There is also an effort to reduce the overhead of scheduling by managing queues such that selection is O(1) instead of O(n) [10]. Another technique to reduce scheduling latency is to increase system clock frequencies [11]. The processing power of modern systems may tolerate the interrupt overhead [12].…”

Section: Related Workmentioning

confidence: 99%