# Thread: User Modeling Scenario

1. ## User Modeling Scenario

Any input on my calculations below would be greatly appreciated. Is there a good forum to go to to discuss user modeling? Or am I in the right place?

I am trying to calculate the number of active sessions during peak hour in production.

BTW...no, I can't pull active sessions from production app servers (long story). The data below is from a tool that tracks user interactions with the web site.

The story is the same but the numbers have been changed to protect the innocent...

- Peak hour of this year we had 31,463 Visits to our site
- Average visit lasted 10:33 minutes
- Our app server timeout is 10 minutes
- So, average session life is 20:33 minutes

Based on these facts I calculated…
31,463 visits / 60 min = 524.4 vis/min

Assuming that during peak hour sessions will timeout as fast as they are created and each visitor creates 1 session per visit…
524.4 vis/min * 20.5 min = 10750 visits which is equivalent to 10,750 sessions exist during any given time during peak traffic.

2. ## Re: User Modeling Scenario

Do any of the users end session by log out vs. time out?

3. ## Re: User Modeling Scenario

no, no login. good question though.

4. ## Re: User Modeling Scenario

Anyone out there use WebSideStory hbx? aka Hitbox.

www.websidestory.com

I am using it to create my User Model. All the above info was gathered using that tool. There is a lot of data.

The reason I ask is I am trying to put together a "dashboard" with the things I need to put together the model.

5. ## Re: User Modeling Scenario

Based on your assumptions, that calculation is valid for an "average peak" load. Depending on the variance from min to min (or in this case, more like 10 min to 10 min) this could either be pretty close to the "instantaneous peak" or a rather low estimate.

If I were to assume that these numbers accounted for all of the available data, I'd recommend a "peak factor" of 1.3 to 1.5 times the average of 10,750 that you calculated. For easy math, I'd use 15,000 as my "estimated peak" load for testing purposes and for my simulated peak hour, I'd have the load rise and fall from about 8,000 to 15,000 in a pattern that ends up matching the 31+k over the course of the test run.

Actually, for testing purposes, I'd go 10k -&gt; 15k that averaged out to about 35k over an hour under the "better safe than sorry" model.

With more data, there are certainly methods to add accuracy, but my take on your question was that this was all the data available.

6. ## Re: User Modeling Scenario

Thanks for your input Scott!

I don't think I can get more granular with the data. Avg over an hour is the smallest. I did plan to scale up the peak to account for peak minute. The fluctuation is a very interesting idea. I will think about how to make that happen...

7. ## Re: User Modeling Scenario

Pleased to be of service. [img]images/icons/wink.gif[/img]

#### 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 11:27 AM.