Hi Raghul,
Thanks for the detailed reply. The overall setup is much clearer now but the issue raised is still illusive.
It is unclear why hits per seconds are identical while throughput is lower in TruClient. Can you share the results with HPE expert via support ticket? You can do so only directly to me, just send me a private message so I can share my business email address
Regards,
Shlomi
Hi Oren,
I am facing some difficulties when trying to upload the zip file. Actually the zip file size is 24.5 mb with all files. Due to that not able to send you through email. In that blog I am not able to upload any file other than jpeg,gif,png. So plz suggest what should I do? Is there any other option ?
Thanks,
AbhijitK
please send me any email to og@hpe.com .
and we will find any solution to make the script minimize
Oren
Please verify the following:
1. in the response snapshot of the step, such a string exists.
2. the web_reg_find is configured to search in the right place, i.e. the target place to search matches the place in which the string is sent by the server
Hi,
Anyone have a reply for this? I would like to know if there is any way to count the number of tabs that are open in firefox?
Any solution for this?
Thanks a lot for the suggestion. I tried the HTTP Protocol for the application I am testing, but the Run Time Viewer says "Javascript is not anabled in the browser", which forced me to use TruClient Protocol. How do I include the performance counters that is mentioned in the link. Is it at the script level or Controller level?
Also, I already have 2 LGs set up for the load test. To add another LG(Windows server 2008), do I just install LoadGenerator.exe on that machine and add it via controller? Or do I need to install a patch on that LG machine. I am using version 12.02.
Hi Raghul,
Just realized I had a mistake in previous message, I meant to say you can open support ticket or send me a private message.
In any case since you already opened a ticket I am sure the support engineers will be able to explain the behavior as well
Regards,
Shlomi
Hi,
1. Web (HTTP/HTML) run time viewer is a naive implementation of browser control which does not support rendering modern applications. It does not indicate anything about the protocol's capability to simulate the right workload against your application assuming you properly handled all dynamic data in your script.
Web(HTTP/HTML) simulate the traffic in the HTTP level. The script is more challenging to create but it is much more scalable.
Bottom line: do not let the Run Time Viewer determine whether or not you use it.
2. If you wish to monitor the Load Generator using performance counters add the Windows Resources graph in Controller's run tab, add the Load Generator machine to be monitored by right clicking the graph and use the context menu to drive this configuration
3. In order to use another Load Generator you need to install the Load Generator sofwtare on that machine and add the Load Generator from the Load Generators window in Controller.
I suggest you go over the introduction section in this online help page: http://lrhelp.saas.hpe.com/en/12.53/help/WebHelp/Content/Controller/c_designing_scenarios_overview.htm
Good Luck!
Shlomi
Hi,
There is no built-in function for that.
Here is a workaround to get this number. I tested it on Firefox and TruClient IE.
var numOfTabs = 0; // Initialize the number of tabs to zero in Evaluate JavaScript step
For (var i=1; i < 100; i++){ // Use TruClient For step from the Toolbox
Activate tab i // Use Generic Browser Action step and change the action to Activate tab in the i place
Catch Any error // Use Catch step to capture any error
Break // The catch step will be triggered when TruClient will try to activate a tab that does not exist
numOfTabs=i; // Evaluate Javascript step again. If we reached this point it means that activate tab succeeded and we can count it for later use
}If use this workaround you will need to activate a specifc tab after this loop in order to ensure the script continues to run successfully
Regards,
Shlomi
Thanks a lot.
I was able to install counters and add more LGs to the controller. However, I am noticing something weird while using Web Protocol. The application that I am using is having lot of dynamic objects and is evident from the fact that when I use TruClient, I have to apply Javascript to almost every object in the app to be able to identify them when i replay the script. But, If I develop the same use case in Web Protocol, the replay seems to be work without errors. I was expecting it to throw errors on the objects that are dynamic, but it didn't and it works everytime when I replay it.
The design studio feature shows some values that need to be correlated foreg:- the login button, but when I do the correlation, the script throws error saying it couldn't identify the object in the parameter and If i remove the correlation, it works fine. I am not able to understand the behavior. Can you please help ?
Hi,
The hardware is be fine. Both are physical machines , quad cores with 8GB RAM running Windows Server 2012 R2. This issue happens with only 1 virtual user running a simple Web HTML test where CPU usage does not go above 15% and memory usage is below 50%.
Web (HTTP/HTML) and TruClient are very different in nature. You will not see the same behavior or challenges with them. There is no room for comparison.
I suggest you to follow LoadRunner Tutorial in order to learn the basics. it is impossible to instruct you over the community since the quesiton is relatively basic and very general. I suggest focusing on the first 4 lessons for a start
Thanks for the response. I understand the differences between both the protocols, but the question is not coming from understanding their nature. I have been handed over some scripts from a 3rd party that's used the same LR version I am using (12.02) and I saw their scripts had parameterized the buttons and text boxes that are dynamic. Those scripts are developed on Web(HTTP/HTML) Protocol as well. I am only using them to identify the use cases and develop LR script that can run in our network. I just have to replace few URLs in the script in order to re-use it, but it won't run and would throw errors related to parameterization. Now, if i remove the correlation in those scripts, it would work. I am trying to understand why there is such a difference even when both the parties are using same protocols that's built on the same framework, just different URLs.
this turned out to be an internal desktop configuration issue, not an HP/truclient problem. the firefox folder within the loadgenerator installation folder had been mistakenly removed.
Sorry for misjudging the situation.
Can it be that you are working with two different versions? Maybe the script was created in a newer version and you are trying to edit in an older one?
It is very hard to troubleshoot it without actually seeing it. Sorry
Hi,
I am working on a scenario where i need to delete record from listbox using trueclient portocol in vugen, while recording, the scirpt captures the entire text of the record to be deleted form listbox. how can i select a new recrod during reply, I am working on true client for first time, pls can anyone help me with it.
Hi,
From the screenshot you shared it seems like you need to select a row in the table based on data captured earlier in the script or based on data from a parameter file. Once the correct row will be selected you can add a step that will click on remove.
It is not easy to provide exact steps that will fit your case but I can direct you to example of the practices you should use to accomplish it
As a beginner in TruClient I really suggest watching these webinars
Regards,
Shlomi
Hi,
I just updated my community edition from 12.02 to 12.53 also updated windows os from Win7 to Win10. Now when I start recording a scenario it does not invoke firefox browser.
Please advice.
Thanks,
Mubbashir.