Scribd
Upload
115 views13 pages

Borrowed Virtual Time Scheduling

Borrowed Virtual Time (BVT) Scheduling is designed to support low-latency dispatching for real-time applications while allowing weighted CPU sharing among various application classes. The scheduler uses effective virtual time for dispatching threads, enabling latency-sensitive threads to borrow virtual time from future allocations. BVT efficiently balances real-time constraints without disrupting long-term CPU sharing, making it suitable for both hard and soft real-time applications.

Uploaded by

23p1192
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
115 views13 pages

Borrowed Virtual Time Scheduling

Borrowed Virtual Time (BVT) Scheduling is designed to support low-latency dispatching for real-time applications while allowing weighted CPU sharing among various application classes. The scheduler uses effective virtual time for dispatching threads, enabling latency-sensitive threads to borrow virtual time from future allocations. BVT efficiently balances real-time constraints without disrupting long-term CPU sharing, making it suitable for both hard and soft real-time applications.

Uploaded by

23p1192
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 13

BORROWED VIRTUAL TIME

SCHEDULING
PRESENTED BY:
SAGAR S TALAGATTI – 2GI23SCS08
OUTLINE
• Introduction To Borrowed Virtual Time Scheduling
• Example
• Conclusion
• References
INTRODUCTION TO BORROWED VIRTUAL TIME
SCHEDULING
• Objective: To support low-latency dispatching of real-time applications as well as a weighted
sharing of the CPU among several classes of applications.
• Supports scheduling of a mix of applications, some with hard, some with soft real-time constraints,
and applications demanding only a best effort.
• Each thread i has:
1. Effective virtual time Ei
2. Actual virtual time Ai
3. Virtual time warp Wi

• The scheduler thread maintains its own scheduler virtual time (SVT), defined as the minimum
actual virtual time Aj of any thread.
• The threads are dispatched in the order of their effective virtual time, Ei, using a policy called the
earliest virtual time (EVT).
INTRODUCTION TO BORROWED VIRTUAL TIME
SCHEDULING
• The virtual time warp allows a thread to acquire an earlier effective virtual time.
• That is, to borrow virtual time from its future CPU allocation.
• The virtual warp time is enabled when the variable warpBack is set.
• In this case a latency-sensitive thread gains dispatching preference as:

• The algorithm measures the time in minimum charging units (mcu) and uses a time
quantum called context switch allowance (C).
• C measures the real time a thread is allowed to run when competing with other threads,
measured in multiples of mcu.
INTRODUCTION TO BORROWED VIRTUAL TIME
SCHEDULING
• Typical values for the two quantities are mcu = 100 μsec and C = 100 msec.
• A thread is charged an integer number of mcu.
• Context switches are triggered by traditional events:
• the running thread is blocked waiting for an event to occur
• the time quantum expires
• an interrupt occurs
• a thread becomes runnable after sleeping

• Thread τi becomes runnable after sleeping, its actual virtual time is updated as
follows:
INTRODUCTION TO BORROWED VIRTUAL TIME
SCHEDULING
• If there are no interrupts, threads are allowed to run for the same amount of virtual
time.
• Threads have weights, a thread with a larger weight consumes its virtual time more
slowly.
• Each thread τi maintains a constant ki and uses its weight wi to compute the amount
used to advance its actual virtual time upon completion of a run:
Ai ← Ai + ∆
• Given two threads a and b: ∆ = ka / wa = kb / wb

• The EVT policy requires that every time the actual virtual time is updated, a context
switch from the current running thread τi to a thread τj occurs if:
EXAMPLE
EXAMPLE
EXAMPLE
EXAMPLE
• Table 6.5 shows the effective virtual time of the two threads at the time of each context switch.
• At that moment, its actual virtual time is incremented by an amount equal to ∆ if the thread
was allowed to run for its time allocation.
• The scheduler compares the effective virtual time of the threads and first runs the one with the
minimum effective virtual time.
• When a thread is running, its effective virtual time increases as the real time increases; a
running thread appears as a diagonal line.
• When a thread is runnable but not running, its effective virtual time is constant. A runnable
period is displayed as a horizontal line.
• We see that the two threads are allocated equal amounts of virtual time, but thread a, with a
larger weight, consumes its real time more slowly.
CONCLUSION
• BVT is an efficient algorithm for low-latency dispatching of real-time applications.
• It allows for weighted sharing of the CPU among several classes of applications.
• It allows each thread to borrow its own future CPU allocation and hence doesn’t
disrupt long-term CPU sharing.
REFERENCES
• Cloud Computing Theory and Practice by Dan C. Marinescu, Morgan Kaufmann
Publications
THANK YOU

You might also like