Can I get Vuser Foot prints documentation i.e for How many users can have a single LG machine.
What I am trying to say is let's say for ex: FLEX protocol : How many users can handle single LG?
HP has stopped providing the footprint estimate document that was created back in LoadRunner 8.X because it was not accurate. We learned that to truly understand how many Vusers a generator could handle you had to consider far to many factors. Some of those factors are: The amount of free memory available on the generator, the OS of the generator, what other software was running on the generator, the CPU speed, the number of CPU's, how much of the software already running on the generator could support multiple CPUs, the disk speed, how full the disk is, the network speed, how full the network is, the protocol of the script, any custom code in the script, etc, etc, etc.
To determine how many virtual users your individual generator can handle complete the following steps.
1) RDP to the load generator
2) After the entire boot cycle is complete open Task Manager, switch to the performance tab and note the amount of Free Physical Memory (FPM). Please note FPM is not the amount of RAM installed on the server, it is the amount of free ram after the OS and all other software is loaded.
2.1) Multiple FPM by 80% to get your Available Free Physical Memory (AFPM).
3) Set up a 1 user load test using LR controller with your script setup to run for 5 minutes on the Generator you are reviewing (make sure it isn't going to localhost)
4) Run the test and note how much the FPM drops. This is the amount of RAM consumed by the 1st vuser (1RAM)
5) Reconfigure the load test to run 2 users. Have the first user start and 5 minutes later have the second user start and then run both users for 5 minutes
6) Run the test and note how much FPM drops when the second user kicks off and is running. This is the amount of RAM consumed by all subsequent vusers (XRAM)
7) Now take ((AFPM - 1RAM ) \ XRAM) 1 and this is how many virtual users that generator can run of that script
Here is an example with numbers:
- You determine that your FPM is 1024mb (1GB)
- 1024 * .8 = 819mb. This is your AFPM
- Your first load test shows a drop of 32mb of FPM (1RAM)
- Your second test shows a drop of 4MB of FPM when the second user is fully running (XRAM)
- ((AFPM -1RAM)\XRAM) 1 or in other words, ((819 - 32) \ 4) 1 = 197 (because you should always round down any decimal)
Also don't forget that many GUI based protocols (Citrix, RDP, and most protocols that have the term GUI in them) can further be limited by GDI resources. GDI resources are limits of the graphics card and regardless of how much FPM your computer has, once GDI resources are consumed, no additional users will run. There is no magic formula for estimating GDI and the only way to increase GDI is to change video cards on the computer.
I hope this helps you with the information that you seek