Modeling and measuring multiprogramming and system overheads on a shared-memory multiprocessor: Case study

Dimpsey, R. T.; Iyer, Ravishankar K.

doi:10.1016/0743-7315(91)90009-x

“…This may limit the realistic clock interrupt rate to 1000 Hz or a bit more, but probably not as high as 5000 Hz (in this case the P-IV is substantially better than the other platforms, but this is due to using performance relative to 100 Hz, which was worse than for other platforms for an unknown reason). Note, however, that 1000 Hz is an order of magnitude above what is common today, and already leads to significant benefits, as shown in subsequent sections; the added overhead in this case is just a few percentage points, much less than the 10-30% which were the norm a mere decade ago [7]. Our measurements also allow for an assessment of the relative costs of direct and indirect overhead.…”

Section: Clock Resolution and Overheadsmentioning

confidence: 75%

Effects of clock resolution on the scheduling of interactive and soft real-time processes

Etsion

¹

,

Tsafrir

²

,

Feitelson

³

2003

Proceedings of the 2003 ACM SIGMETRICS International Conference on Measurement and Modeling of Computer Systems

View full text Add to dashboard Cite

It is commonly agreed that scheduling mechanisms in general purpose operating systems do not provide adequate support for modern interactive applications, notably multimedia applications. The common solution to this problem is to devise specialized scheduling mechanisms that take the specific needs of such applications into account. A much simpler alternative is to better tune existing systems. In particular, we show that conventional scheduling algorithms typically only have little and possibly misleading information regarding the CPU usage of processes, because increasing CPU rates have caused the common 100 Hz clock interrupt rate to be coarser than most application time quanta. We therefore conduct an experimental analysis of what happens if this rate is significantly increased. Results indicate that much higher clock interrupt rates are possible with acceptable overheads, and lead to much better information. In addition we show that increasing the clock rate can provide a measure of support for soft real-time requirements, even when using a general-purpose operating system. For example, we achieve a sub-millisecond latency under heavily loaded conditions.

show abstract

“…This may limit the realistic clock interrupt rate to 1000 Hz or a bit more, but probably not as high as 5000 Hz (in this case the P-IV is substantially better than the other platforms, but this is due to using performance relative to 100 Hz, which was worse than for other platforms for an unknown reason). Note, however, that 1000 Hz is an order of magnitude above what is common today, and already leads to significant benefits, as shown in subsequent sections; the added overhead in this case is just a few percentage points, much less than the 10-30% which were the norm a mere decade ago [7].…”

Section: Clock Resolution and Overheadsmentioning

confidence: 99%

Measurement of Performance

Morchin¹,

Oman²

2011

Electric Bicycles

0

View full text Add to dashboard Cite

It is commonly agreed that scheduling mechanisms in general purpose operating systems do not provide adequate support for modern interactive applications, notably multimedia applications. The common solution to this problem is to devise specialized scheduling mechanisms that take the specific needs of such applications into account. A much simpler alternative is to better tune existing systems. In particular, we show that conventional scheduling algorithms typically only have little and possibly misleading information regarding the CPU usage of processes, because increasing CPU rates have caused the common 100 Hz clock interrupt rate to be coarser than most application time quanta. We therefore conduct an experimental analysis of what happens if this rate is significantly increased. Results indicate that much higher clock interrupt rates are possible with acceptable overheads, and lead to much better information. In addition we show that increasing the clock rate can provide a measure of support for soft real-time requirements, even when using a general-purpose operating system. For example, we achieve a sub-millisecond latency under heavily loaded conditions.

show abstract

“…Since 0 is the probability the resource is free, the resource utilization is given by U = 1 ? 0 (5) The average number of processors in the context switching state is given by the equation…”

Section: Thenmentioning

confidence: 99%