Page 89 - DCAP608_REAL TIME SYSTEMS
P. 89

Real Time Systems




                    Notes            violated in any situation, an offline approach only needs to show that one situation exists,
                                     i.e., the scheduling table, in which the timing constraints are met: instead of a “for all”
                                     proof, considering even situations which may never occur during the runtime of a system,
                                     a “there exists one” suffices.
                                   Thus, the process follows the steps: task set  with timing  constraints-schedulability test  and
                                   determination of rules (for example, via directives priority or deadline) – execution of rules by
                                   runtime scheduler-timing constraints met.

                                   Looking closer, we can see that TT real-time scheduling work in the same way. Instead of a
                                   definition of rules, e.g.,  earliest deadline  first, “the decisions on  which task  to execute  are
                                   represented in the scheduling table, schedule next task as given table”.
                                   While TT scheduling has to assume a periodic world and ET provides flexibility for tasks with
                                   not fully  known parameters,  example, aperiodic,  the difference  concerns  mostly  runtime
                                   execution without guarantees. When offline guarantees are required, task parameters have to be
                                   known offline: without worst case execution, period  or maximum arrival frequency, offline
                                   guarantees cannot be given, independent of the scheduling paradigm used.
                                   Hence, we can conclude that the terms “offline” and “online” scheduling cannot be seen  as
                                   disjoint in general. Real-time scheduling requires offline guarantees, which require assumptions
                                   about online behavior at design time. At runtime, both offline and online execute according to
                                   some (explicitly or implicitly) defined rules, which guarantee feasibility. The question “offline”
                                   vs. “online” is thus less black and white, but more about how much of the decision process is
                                   Thus, both offline and online are based on a substantial offline part. The question is then where
                                   to set the tradeoff between determinism – all decisions offline - and flexibility – some decisions
                                   online.




                                     Lab Exercise Search through internet and make points of differences between the terms
                                     “offline” and “online” scheduling.

                                   Self Assessment


                                   State whether the following statements are True or False:
                                   7.  Defined rules are represented as priorities.
                                   8.  An offline schedulability test can be used to show that, if a set of rules is applied to a given
                                       task set at runtime, all tasks will meet their deadlines.
                                   9.  Major representative lines of  such algorithms  are based  on fixed  priorities, e.g.,  rate
                                       monotonic or dynamic priorities.

                                   10.  The scheduling table of offline scheduling provides a “proof by construction” that all
                                       timing constraints will be met.
                                   11.  When offline guarantees are required, task parameters have to be known online.

                                   12.  The terms “offline” and “online” scheduling cannot be seen as disjoint in general.












          84                                LOVELY PROFESSIONAL UNIVERSITY
   84   85   86   87   88   89   90   91   92   93   94