For this assignment you will be trying to resolve a series of possible computer problems and determine what information you would gather to diagnose the problem and then solve the problem. To avoid extreme complexity we will assume that we are addressing issues on a single server that has directly attached storage. That is, we are not dealing with blade servers that are running in virtual mode that store data in a network attached storage array. While the latter is the standard mode of operation in many server environments, this adds too much complexity to the following problems.
For this assignment I will allow you to work with a partner so that you can bounce ideas off of each other. However, if you wish to work alone, that is your option. If you work with a partner, your team will submit only one document. Be sure that you place both names on your document.
In the first part of the assignment you will be given a situation that has the potential to cause poor system performance and you will discuss the possible behavior the system could probably exhibit.
Situation 1: A server has a series of disk drives attached to it and a series of different applications run on the server. In an attempt to reduce I/O contention between applications, groups of applications are assigned to a single disk drive. One of the disk drives is badly fragmented, however the remaining disk drives have been recently reorganized (i.e. fragmentation primarily eliminated). Describe what the result of this situation could possibly be with respect to the applications being run. What additional measures (other than disk fragmentation) would be useful in determining that only the fragmentation is causing the problems? The solution to the problem is obvious, but discuss the problems there could be when trying to solve the problem.
Situation 2: A server has the following statistics: processor busy 60%, operating system active 48% and application systems 12% running, real memory 100% utilization, paging activity – very high, I/O very busy – dominated by page pool activity. Discuss the possible problems with respect to the programs running that could be occurring on this system. What additional measures would you want to see to determine exactly what was causing this performance?
For the following situations there are problems occurring with a system and you are to first conjecture what the possible causes could be for the problems. Then you are to discuss what system measurements you would like to see and explain how these measurements would help you to solve the problem.
Situation 3: You are using the SAP ERP system and your response time has becomeunacceptably slow. You ask workers in cubes near you if they are also having response time problems and they say that they are. Then you ask if they are also using the SAP ERP system and some that are experiencing problems say yes and other say no. You call the help desk and they say that only the office where you work is reporting problems and no other offices for the company are having performance problems. Assist the help desk personnel by suggesting what to monitor and give them some suggestions of the root of the problem.
Situation 4: The same situation as situation 3 above except when you call the help desk, they say that everyone using the server you are using is having poor response time. Again, assist the help desk personnel by suggesting what to monitor and give them some suggestions of the root of the problem.
This assignment will be graded on both the quality of your written expression and the content of your answers. This assignment is limited to a maximum of four pages. You must submit an electronic copy of your assignment via Vista and a paper copy in class by Wednesday October 12, 2011.