scheduler-simulator

git clone git://git.codymlewis.com/scheduler-simulator.git
Log | Files | Refs | README

Report.md (3842B)


      1 ---
      2 title: Comp2240 - Assignment 1
      3 author: Cody Lewis [c3283349@uon.edu.au](mailto:c3283349@uon.edu.au)
      4 date: \today
      5 geometry: margin=2cm
      6 linkcolor: blue
      7 ---
      8 
      9 # Datafile 1
     10 
     11 The Narrow Round Robin scheduling algorithm was the best performing for the
     12 first data file, and First Come First Served performed the worst. Both Round
     13 Robin and Feedback performed equally in this case. The performance of these
     14 algorithms is maximised by the reduction of time spent of processes for longer
     15 amounts of times spent on processes in the case where all the processes arrive
     16 at the same time. This may be observed with First Come First Served perfoming
     17 the worst, since it sequentially processes each whole process. Next, Round
     18 Robin and Feedback which perform equally as they both have a constant slice
     19 size, yet they perform better than FCFS as involve more switching between
     20 processes which allows smaller processes to wait less before completing.
     21 Finally Narrow Round Robin is the best performing as it reduces the slice size
     22 for each time it processes a process, this makes it so that less expensive
     23 processes have to wait for less time before their turnaround.
     24 
     25 # Datafile 2
     26 
     27 In this data file, the processes each come in a differing times, this made the
     28 scheduling algorithms in order from worst performing to best performing
     29 Narrow Round Robin, Round Robin, First Come First Served, and Feedback. Both
     30 Round Robins performed badly in this case as they switch through processes on
     31 a queue structure with no priority system, this means the that the first
     32 process will process for a slice, other processes get added in, then when the
     33 slice is finished, the first will have to wait through the slices of the added
     34 processes before continuing. Narrowing makes this worse by making so that first
     35 process has even less processing time before switching. First Come First Served
     36 performs a bit better as it completes each whole process when it is dequeued.
     37 Feedback Perfroms the best in this case due to the priority system, this means
     38 that even though there are slices, each process is placed on a tier that
     39 marks the immediacy of the need to process them, thus making for less waiting
     40 times.
     41 
     42 # General Observation
     43 
     44 From these observations, it can be concluded that the Feedback scheduler is on
     45 average the best performing of the scheduling algorithms, and that whether the
     46 Round Robin schedulers or the First Come First Served schedulers perform better
     47 then one another is dependant on the timing of how the processes arrive.
     48 
     49 The Feedback scheduler performs best on average due to its combination of
     50 slicing processing and the priority system. The slicing of processing makes it
     51 so that a process does not take up the processor for too long of periods of
     52 time, and thus allows shorter processes to complete without having to wait too
     53 long for those longer ones. The priority system makes it so that the a service
     54 does not have to wait to long to start processing, this allows shorter
     55 processes to be finished quite quickly, while longer ones may have spend a bit
     56 of time waiting, although not too long as the shorter processes will be
     57 finished and eliminated from the queue quite quickly.
     58 
     59 The First Come First Served scheduler performed better when processes arrive at
     60 different times as it processes those sequentially in order of arrival, this
     61 reduces waiting time in a lot of cases.
     62 
     63 The Round Robin schedulers performed best when all of the processes arrived
     64 at the same time. This is due to the slicing of processing, which means that
     65 smaller processes do not have to wait through some of the bigger ones to
     66 complete processing before having a chance to process. The narrowing variation
     67 of the scheduler tended to perform better as it tended to give the processes
     68 more of a chance to do some processing in a similar way to that of priority
     69 system.