Real Time Task Scheduling

ͽ Basic concepts related to scheduling

ͽ 1. Valid schedule
ͽ 2. Feasible schedule
ͽ 3. Scheduling points
ͽ 4. Preemptive scheduler
ͽ 5. Optimal scheduler
ͽ 6. Proficient scheduler
ͽ 7. Jitter
ͽ 8. Utilization
 Classification of real-time scheduling
algorithms based on scheduling points

Clock Driven Event driven Hybrid

• Table-driven • Simple priority-based • Round-robin
• Cyclic • Rate Monotonic Analysis
(RMA)
• Earliest Deadline First (EDF)
 Classification of real-time scheduling
algorithms based on acceptance test

Planning-based Best effort

 Classification of real-time scheduling
algorithms based on target platform

Uniprocesso Multiproces
Distributed
r sor
 Clock-Driven Scheduling
ͽ Scheduling points are determined by timer interrupts.
ͽ Scheduling decisions regarding which task to run next.
ͽ Also called off-line schedulers because these schedulers fix the schedule
before the system starts to run.
ͽ Adv: incur very little run time overhead
ͽ Disadv: can not satisfactorily handle aperiodic and sporadic tasks since the
exact time of occurrence of these tasks can not be predicted. (also called a
static scheduler)
ͽ Two important clock-driven schedulers:
ͽ 1. Table-driven and
ͽ 2. Cyclic schedulers.
 1. Table-Driven Scheduling
ͽ precompute which task would run when
ͽ store the schedule in a table at the time the system is designed or configured.
ͽ The application programmer can be given the freedom to select his own
schedule for the set of tasks in the application and store the schedule in a
table.
 1. Table-Driven Scheduling
ͽ major cycle / hyper period of the set of tasks ST
ͽ LCM(p1, p2, .., pn) is called the major cycle of the set of tasks ST.
ͽ Determines the size of the schedule table.

ͽ A major cycle of a set of tasks is an interval of time on the time line such
that in each major cycle, the different tasks recur identically.

ͽ For any given task set it is sufficient to store entries only for LCM(p1,
p2, .., pn) duration in the schedule table.
 Table-Driven Scheduling

Task p e

1. T1
T2
20
100
5
20
T3 250 30

Task P e
2. T1 40 1
T2 10 2
T3 20 1
T4 20 5
 Table-Driven Scheduling
Task P e
T1 4 1
T2 5 1.8
T3 20 1
T4 20 2

ͽ Utilization: 0.25, 0.36, 0.05, and 0.1

ͽ Example S
 2. Cyclic Schedulers
ͽ Popular and extensively used in the industry.
ͽ A large majority of all small embedded applications being manufactured
presently are based on cyclic schedulers.
ͽ Simple, efficient, and are easy to program.
ͽ Example: temperature controller. A temperature controller periodically
samples the temperature of a room and maintains it at a preset value. Such
temperature controllers are embedded in typical computer-controlled air
conditioners.
 2. Cyclic Schedulers
ͽ repeats a precomputed schedule.
ͽ The precomputed schedule needs to be stored only for one major cycle.
ͽ Each task in the task set to be scheduled repeats identically in every major
cycle.
ͽ The major cycle is divided into one or more minor cycles.
ͽ Each minor cycle is also called a frame.
ͽ scheduling points occur at frame boundaries.
ͽ This means that a task can start executing only at the beginning of a frame.
 2. Cyclic Schedulers
ͽ The size of the frame to be used by the scheduler is an important design
parameter and needs to be chosen very carefully.
ͽ A selected frame size should satisfy the following three constraints.
ͽ 1. Minimum Context Switching
ͽ 2. Minimization of Table Size
ͽ 3. Satisfaction of Task Deadline
 2. Cyclic Schedulers
ͽ 1. Minimum Context Switching
ͽ This constraint is imposed to minimize the number of context switches occurring
during task execution.
ͽ task instance must complete running within its assigned frame.
ͽ Unless a task completes within its allocated frame, the task might have to be
suspended and restarted in a later frame.
ͽ This would require a context switch involving some processing overhead.
ͽ To avoid unnecessary context switches, the selected frame size should be larger
than the execution time of each task, so that when a task starts at a frame
boundary it should be able to complete within the same frame. Formally, we can
state this constraint as: max({ei}) < F where e is the execution times of the of
task T, and F is the frame size.
ͽ This constraint imposes a lower-bound on frame size.
 2. Cyclic Schedulers
ͽ 2. Minimization of Table Size.
ͽ This constraint requires that the number of entries in the schedule table should be
minimum in order to minimize the storage requirement of the schedule table.
ͽ Remember that cyclic schedulers are used in small embedded applications with
very small storage capacity.
ͽ So, this constraint is important to the commercial success of a product.
ͽ Minimization of the number of entries to be stored in the schedule table can be
achieved when the minor cycle squarely divides the major cycle.
ͽ When the minor cycle squarely divides the major cycle, the major cycle contains an
integral number of minor cycles (no fractional minor cycles).
ͽ We can formulate this constraint as:
 2. Cyclic Schedulers
ͽ 3. Satisfaction of Task Deadline.
ͽ This third constraint on frame size is necessary to meet the task deadlines.
ͽ This constraint imposes that between the arrival of a task and its deadline, there
must exist at least one full frame.
ͽ This constraint is necessary since a task should not miss its deadline.
ͽ Consider this: a task can only be taken up for scheduling at the start of a frame.
 2. Cyclic Schedulers
ͽ If between the arrival and completion of a task not even one frame exists, a situation
as shown in Fig. 7 might arise. In this case, the task arrives a little after the kth
frame has started. Obviously it can not be taken up for scheduling in the kth frame
and can only be taken up in the (k+1)th frame. But, then it may be too late to meet
its deadline since the execution time of a task can be upto the size of a full frame.
This might result in the task missing its deadline since the task might complete only
at the end of (k+1)th frame.

ͽ This constraint imposes a upper-bound on frame size.

Task P e
T1 4 1
T2 5 1
T3 20 1
T4 20 2

Task P e
T1 4 1
T2 5 2
T3 20 5
 Find LCM
Task P e
T1 4 1 1. max({ei}) < F
T2 5 2
T3 20 5 2.

3.

Task P e
T1 4 1
T2 5 2
T3 20 1
T3 20 2
T3 20 2
 Cyclic Schedulers vs Table-driven Schedulers
ͽ A cyclic scheduler needs to set a periodic timer only once at the application
initialization time. This timer continues to give an interrupt exactly at every frame
boundary.
ͽ But in table-driven scheduling, a timer has to be set every time a task starts to run.
The execution time of a typical real-time task is usually of the order of a few
milliseconds. Therefore, a call to a timer is made every few mill Seconds.
ͽ This represents a significant overhead and results in degraded system performance.
Therefore, a cyclic scheduler is more efficient than a table-driven scheduler.
ͽ This probably is a reason why cyclic schedulers are so overwhelmingly popular
especially in embedded applications.
ͽ However, if the overhead of setting a timer can be ignored, a table-driven
scheduler is more proficient than a cyclic scheduler because the size of the frame
that needs to be chosen should be at least as long as the size of the largest execution
time of a task in the task set.