CS 5113 and 4113 Fall 20

Logo

This is the web page for Fall 2020 Operation Systems Theory at the University of Oklahoma.

View the Project on GitHub oudalab/cs5113fa20

Assignment 1: Understanding Global Latency

(Due September 15)

CS (4|5)113 Fall 2020

Introduction

Latency is described as the length of time a signal takes to travel between two components. Famously, on the web is a chart that is described as Latency Numbers Everyone Should Know. The document describes the astronomical differences in latency times between machine components. In this assignment, you will get a a better understanding of latency in systems. Your task is to perform latency testing between servers on different continents.

Measuring Latency

There are various tools to measure latency between computing element. To compute the time between systems on teh internet we use round-trip delay or round-trip time (RTT) time. This is the time it takes a message to be sent to a system plus the time it take the system to reply. Note, when measuring latency in single node systems, the user is typically able to control the source and target destinations so one way latency time is typically used. We can use the traceroute (tracert on windows) command to measure the time between two nodes. The traceroute command line tool sends a packet and retrieves the acknowledgement of each hop on the path of a messages way to its destination. This blog post give a good description of how to read a traceroute.

Assignment

For this assignment, compute RTT time between continents. You will need to find source and destination location to perform the timings. Undergrad students must compute RTT times between North America, Europe, and Asia. Graduate students must compute RTT times between North America, Europe, Asia, Australia, and Africa. One method is to create machines in your Google cloud accounton each continent. You may also find a server on the internet that is in the city and continent you desire. Note, graduate students will have to find a machine located in Africa because Google does not currently have a data center on the continent.

Submission

Create a table with the headings below. Perform an all-pairs comparison of a city in each continent. Undergrad students need to choose a city in North America, Europe, Asia. Gradduate students need to choose a city in North America, Europe, Asia, Australia, and Africa. You only need to compare one way. After creating the table, include in the file a list or screenshot of the commands you used to obtain the RTT times. Submissions will be through canvas and should be as a .pdf file or a text/markdown file.

Source Destination Time Metadata
_ _ _ _
       
       

Be sure to give the source and destination as cities. Also, give the proper units for each time. Under metadata, you should supply all extra information that is necessary to understanding the results. For example, metadata may say that the connections were between two google networks on a virtual private network. As another example, metadata may also include the time of day of the traceroute.

Grading

Task Percentage
All Cities 90 %
Metadata Supplied 10%
  100%

For information on how to create a virtual machine in another region, revisit the previous course lecture. Please give enough information in the meta data colum for someone to reasonable replicate your result.


Go back to CS (4/5)113