Publication detail

Hardware scheduler monitor

ARM, J. BRADÁČ, Z.

Original Title

Hardware scheduler monitor

English Title

Hardware scheduler monitor

Type

conference paper

Language

en

Original Abstract

This paper takes a brief look into fault tolerant techniques and presents a proposal of hardware hard real-time OS scheduler monitor. Purpose of the new proposal is to detect earlier software failures caused by unpredictable hardware faults or system overload. The proposal focuses more precisely on scheduler failures like a deadline miss and a delayed switching context time. Proposed monitor can be used as a standalone scheduler monitor or as a part of redundant system controller. The assumption is that the scheduler monitor detects hardware and RTOS kernel faults earliest. This should enhance RTOS system reliability and safety.

English abstract

This paper takes a brief look into fault tolerant techniques and presents a proposal of hardware hard real-time OS scheduler monitor. Purpose of the new proposal is to detect earlier software failures caused by unpredictable hardware faults or system overload. The proposal focuses more precisely on scheduler failures like a deadline miss and a delayed switching context time. Proposed monitor can be used as a standalone scheduler monitor or as a part of redundant system controller. The assumption is that the scheduler monitor detects hardware and RTOS kernel faults earliest. This should enhance RTOS system reliability and safety.

Keywords

RTOS, scheduler, real-time, monitor, switching context time

RIV year

2015

Released

17.08.2015

ISBN

978-80-214-5239-8

Book

Sborník příspěvků studentské konference Kohútka 2015

Pages from

4

Pages to

7

Pages count

4

BibTex


@inproceedings{BUT115825,
  author="Jakub {Arm} and Zdeněk {Bradáč}",
  title="Hardware scheduler monitor",
  annote="This paper takes a brief look into fault tolerant
techniques and presents a proposal of hardware hard real-time
OS scheduler monitor. Purpose of the new proposal is to detect
earlier software failures caused by unpredictable hardware faults
or system overload. The proposal focuses more precisely on
scheduler failures like a deadline miss and a delayed switching
context time. Proposed monitor can be used as a standalone
scheduler monitor or as a part of redundant system controller.
The assumption is that the scheduler monitor detects hardware
and RTOS kernel faults earliest. This should enhance RTOS
system reliability and safety.",
  booktitle="Sborník příspěvků studentské konference Kohútka 2015",
  chapter="115825",
  howpublished="electronic, physical medium",
  year="2015",
  month="august",
  pages="4--7",
  type="conference paper"
}