User Tag List

Results 1 to 3 of 3
  1. #1
    Junior Member
    Join Date
    Dec 2000
    Tracy, CA U.S.A.
    Post Thanks / Like
    0 Post(s)
    0 Thread(s)

    Things to consider when doing a Load/Stress test using LR


    I am a newbie with doing Load/Stress test in our company but kind of familiar with the LR tool. The part that i am staggering with is things like how to produce a meaningful result. So i have this case study below which i hope i would get input from you on how to setup the performance test.

    1. There is an upgrade on the new payment system for our order processing.
    2. Front end did not change but the architecture of the new payment system changed tremendously
    3. The goal is to come check that the performance for processing the order does not decrease based on the old architecture.

    Things to consider:
    1. Load generator will run on a 800MHZ machine with 512 RAM, using LR 7.02 can have another host with the same specs.

    1. Do we need to consider tha fact that we are stressing the NIC capacity too.
    2. How can we formulate the amount of load that we will put on the system base on the facts that the environment is like production but not similar.
    3. How are we going to relate the result when this load will be performed on a T1 connection.

    Thanks in advance to all your reply.

    Smile :-)
    - zxyvher
    Smile :-)
    - zxyvher

  2. #2
    Join Date
    Aug 2000
    Vancouver, BC, Canada
    Post Thanks / Like
    0 Post(s)
    0 Thread(s)

    Re: Things to consider when doing a Load/Stress test using LR

    Answer to question 2 (out of our Performance Testing Manual):

    Workload characterization involves studying the user and machine environment, observing key characteristics, and developing a workload model that can be used repeatedly.

    Once a workload model is available, the effect of changes in the workload and system can be easily evaluated by changing the parameters of the model. In addition, workload characterization can help you to determine what's normal, prepare a baseline for historical comparison, comply with management reporting, and identify candidates for optimization.

    Typically, you'll find that the alternatives you're asked to evaluate are stated in business- or user-oriented units instead of machine-oriented units. For example, you'll be asked to model the performance impact of adding 150 new extrenal users compared to 150 new users in the main office.

    The advantage of user-oriented units is that the users can easily relate to them and can frequently give a good forecast of future workload. The disadvantage is that they do not automatically relate to system resources used by a workload. As the performance tester/analyst, you must quantify the relationships, turn an increase in the number of users into an increase in total resources used by each workload class. It is not guaranteed that a good relationship exists.

    A workload class is a combination of transactions that have enough in common to be considered one homogeneous workload. The goal is to have complete coverage of system workload with no overlap of classes.

    Depending upon the evaluation technique being used, workloads may be characterized in different forms.
    Some attributes used for classifying a workload include:

    Resource usage
    * Database activity
    * File server activity
    * Server processes

    User Scenarios
    * Business Functions
    * Application type
    * Data entry
    * Reporting
    * Search function
    * Data retrieval
    * Batch program

    Geographical orientation
    * In-house
    * Remote location through T1-link
    * Internet
    * Dialup

    * Front-Office activities (web pages, functionality used while talking on the phone to a client)
    * Back-Office activities (research, printing etc.)

    Organizational unit
    * Workload divided up per group, department, division
    * Workload divided up per role (Executive, Director, Manager, Group lead, worker)
    * Etc.

    At times, it may be appropriate to use multiple criteria to define a single class. Regardless of how you choose to characterize the workload classes, the resulting workload model should capture quantitative information about the real system workload.

    Workload parameters are characteristics of users' requests such as traffic intensity, system load, and resource consumption. Pay special attention to those workload parameters that will be affected by your alternative performance objectives as well as those that have a significant impact on performance - CPU utilization, total I/O operations, memory usage, page faults, and elapsed time.

    By creating a workload model for your old situation that correctly simulates current usage (you can check that by retrieving current production usage stats.), you can implement a simulation on any set of servers and environment.

    If the new system is installed on the same hardware you can use your same workload simulation and compare the before and after. This will give insight in how the new system performs. However, you should strive to test the new system on the production/staging platforms as well.

    Hope this helps.

    Roland Stens

  3. #3
    Join Date
    Aug 2001
    Post Thanks / Like
    0 Post(s)
    1 Thread(s)

    Re: Things to consider when doing a Load/Stress test using LR


    When tuning an application you should not be introducing artificial bottlenecks to distort your load (as in the case of the t1). If you must run the test from a remote location, stage your virtual user hosts as close to the server under examination as possible. Ideally, on the same switch in the same rack.

    Only after you know your application scales should you move your load out across the network to measure the delay added by the network.

    James Pulley
    iTest Solutions, Inc
    704-243-2854 (voice)

    [This message has been edited by jpulley3 (edited 02-27-2002).]
    James Pulley

    Replace ineffective offshore contracts, LoadRunnerByTheHour. Starting @ $19.95/hr USD.

    Put us to the test, skilled expertise is less expensive than you might imagine.

    Twitter: @LoadRunnerBTH @PerfBytes



Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts

vBulletin Optimisation provided by vB Optimise v2.6.0 Beta 4 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
User Alert System provided by Advanced User Tagging v3.0.9 (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Questions / Answers Form provided by vBAnswers (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
vBNominatevBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Feedback Buttons provided by Advanced Post Thanks / Like (Pro) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
Username Changing provided by Username Change (Free) - vBulletin Mods & Addons Copyright © 2016 DragonByte Technologies Ltd.
BetaSoft Inc.
Digital Point modules: Sphinx-based search
All times are GMT -8. The time now is 11:00 PM.

Copyright BetaSoft Inc.