Excellent article!!! LoadRunner is popular automation testing tool used for validating a software application/system under load. It delivers most precise information about the performance, functionality and behavior of the software product. Loadrunner Training in Chennai | Loadrunner training Chennai|Qtp training in Chennai
I haven't face such scenario, so do not have practical experience. I googled and found that there is no such function in SAP GUI which we can use for file download time. Could you please raise a ticket to Microfocus may they have some solutions?
Kindly share if you get any solution, this will help others.
Since Mercury days you can purchase a premium protocol for SAPGUI. It's about double the cost of a standard HTTP license. This is because SAP communicates via RPC rather than standard HTTP. It works like a charm.
Not right now, but I got some requests to document these posts and make them handy. I will try to prepare a handy document and publish it soon (when I get some free time).
Hi Gagan, A question to you. i have a script ran with 100 users on controller by using 2 LGs,once test complete how i can get response times of my scripts based on LG wise? or where i can see those such type of results in analysis?
You can not filter the result on the basis of LGs, because on completion of the test, reports from all the LGs are collected by Controller and prepared a combined report.
Hi Gagan, I have not worked in controller but in performance center one set up exists where we can filter based on hosts.I am not sure it will help or not.
Hello Sir, Your blog is wonderful. I want to ask solution to a scenario. Suppose, there is a scenario where a user searches flight between 2 cities. Now entering those details and clicking Search, he will get many response as multiple flights will be coming. Now he will select any one and book it. Now the scenario is, suppose we are running the above script with 10 users parallely, they will enter cities details to find flight, now what I want is each user selects a unique flight to book, like as, 1st user books 3am flight, 2nd user books 5am flight, 3rd user books 8am flight, 4th user book 10am flight, and like that. So how to achieve this scenario. I.e. how to capture all response and pass them as parameter to each user differently. Hope this is clear. Thanks.
The easiest solution is to get the flight-related info in a parameter file and then use as per your requirement.
Else, the first user will pick the flight-related information using correlation and write them in a file which is shared between all the users. This file file be your parameter file with sequential order setting. When another user come and book the flight he will pick the next flight listed in the file.
Alternate Solution: If you your requirement fulfills then you can also capture the flight details using correlation and then pass randomly to the each user during runtime.
I am posting the answer how this type of scenario can be achieved in JMeter. I have got the solution for this from Gagan Sir's blog itself. The link for the answer is:- https://perfmatrix.blogspot.com/2017/04/apache-jmeter-nested-variable-handling.html Here, what we need to do is, we need to capture all the response values using Regular Expression in a variable and using -1 for matches. This will capture all such vales in that variable. Now, when we debug, we will get all those as variable_1, varibale_2, varibale_3, so on. Now we need to use the above said -V() function and parametrise this 1, 2, 3 according to the Users. The final parameterized function that need to be used where we are passing that response will be like ${__V(variable_${Values}), here ${Values} is for 1, 2, 3, etc passed form CSV file with respect with Users. Now, when we will run this, first user will select first flight, second user will select second flight, third user will select third flight and so on. The above is for JMeter.
However, to achieve the same in Loadrunner (apart from above two solutions i.e. 1) parameterizing from file, and 2) passing randomly), I have asked Gagan Sir, hope he will come up with a similar solution for nested variable handling in Loadrunner.
Every tool has its own way of handling the common things. Like in JMeter dynamic value is handled by RegEx, but in LoadRunner it is primarily handled by correlation function. In LoadRunner, the above scenario can be handled by nesting the function; not the variable.
Timer (as a step) only executes when previous response is arrived completely. In case the request is taking more time to respond then you can increase step download timeout. But it does not have any connection with timer.
May be your case is like this: Once an order is submitted then on the same page 3 states appeared. 1. Initialize 2. In Progress 3. Submitted And you need to wait on the page until submitted state is appeared. In that case you need to write a logic to check the state on the page.
It would be simply text check. Just you need to add web_reg_find() and provide the search for the state (like Submitted). Now you need to iterate this step until 'Submitted' word is found.
Thanks Lot..Very nice information on your blog on Performance testing tool for different tools. I felt something is missing like record settings. Please can you put that one with examples.
Thanks Ramamurthy to highlight this. I posted one article HTML vs URL mode which comes under recording setting. https://perfmatrix.blogspot.com/2017/04/html-mode-vs-url-mode-recording-loadrunner.html
I will cover rest of the setting options in future post.
Hello Sir, I have 2 queries to be asked. 1st.-> Suppose, we have to test an Application, let it be WebTours, for example. So, there, are basically 3 scenarios, first one is login->logout; 2nd one is login->browse flight->logout; and 3rd one is login->browse flight->purchase/book ticket -> logout. Now, suppose while running for 1000 users to do load test, what will be the best approach to achieve the work load model from Controller. Will it be like, record all scenarios in one script with each Scenario as Action, then in Controller, load the same script 3 times, just change the Action Scenario using Action_Block concept, then assign % or number of Users to each script, like 400 users for first Scenario, 300 users for second Scenario, and 300 Users for third Scenario, and then run all the 3 scripts at parallely. (Or we can record 3 scripts each for 3 scenarios and load them in controller and assign number or % of users and run them parallely). Is this the best way to achieve the workload performance run? Please put light on this whether my approach is correct or not, if not correct, Please describe your approach?
Also, please let us know how to achieve the same in JMeter performance test?
Both the methods are correct, but I vote for first method as it is time saving and shows smart work.
Even in JMeter, the 3rd flow will be recorded once in which pages are separated by Transaction controller. Here we need use three thread groups and copy/paste the flow in all three thread groups. The transaction controllers can be disabled/delete where ever not required.
My 2nd Query is:- Suppose we have to do distributed testing and we need to use load generators using JMeter & LoadRunner. So, my question is, to achieve distributed testing, is it necessary to place (or put) the VUser Script or JMX along with CSV data or Parametrization data in Same Place on Slave Load Generators, while doing distributed testing using both JMeter or LoadRunner, or is it like, it is required while doing from JMeter, but not from LoadRunner? Please put a light on this.
In JMeter: 1. Only data files like CSV Data set etc. need to place in the slave and same path needs to be given in the script. In case if you have more than 1 slave then the file must be placed at the same place in all the slaves. 2. No need to place script in the slave.
In LoadRunner: 1. LoadRunner transfers the Parameter files (.dat) to Load Generator before test starts, so no manual intervention is required. 2. If you have added the files (like image/pdf) in the script then again these files are transferred by LoadRunner itself to Load Generator. 3. While creating the script if you have given the absolute path of the file like (C:\LoadRunner\Project\myFile.txt) then you need to place the file in the load generator by creating the same folder structure. 4. No need to place the script at Load Generator, as it is LoadRunner's work.
Could you please clarify whether you are talking about custom module creation in C or you are referring to divide the script flow into different modules for re-use purpose?
Hi Sir, I have one suggestion and few question to ask. 1.) Suggestion:- In Basics tab on this website, the PE Basic will come first and then PT Basic from top. Now, questions:- 1.) Please put some light on both, one, custom module creation in C and how to use them w.r.t LR, and 2nd, dividing the script flow into different modules for re-use purpose, what is this? 2.) In your blog, how you managed to make dropdown for main tabs, any supporting code or setting for this? 3.) What is the major difference between Monitoring tool and Profiling tool with example tools name respectively? 4.) In worst situation, suppose we have get help from LoadRunner officials, so how to do that? In Jmeter, we need to send an email to user@jmeter.org, similarly, how to get help from LR Officials, or Official Community, what is the detailed procedure to write to them and get help?
Please find my point-wise reply: Suggestion: 1. May I know why you are suggesting to swap the position please?
Question: 1. It is as same as dividing the modules in c++ and java. Practically, I do not see anybody who use the modular approach in LoadRunner. Because script recording is easier than customization using modules. Also, if you already have a script and want to use (say login module), then copy the login code and paste in the another script is a good option. This will also make the script fast rather than calling a module. (As per my view only)
2. May I know the reason for asking this question please?
3. Monitoring tool are platform independent and supported (nearly) all the technology. The main purpose of monitoring tool is check the performance of resources like CPU, Memory etc. Example of Monitoring tool is: Micro Focus Site Scope
Profilling tool is specific to technology and helps to identify the performance by running some specific code to measure performance at code level. Example of Profiling Tool: JProfiler
4. You must have SAID (support agreement ID) or Passport ID for that. SAID is provided by LoadRunner once you purchase the licenses. Still, you can try to create an account for Micro Focus and check whether you can able to raise a service request without SAID.
One more Question:- Difference between Session ID, Bearer Access Token, CSRF Token and OAuth 2.0?
As per my work experience, I know 1st two, and writing it here, plz correct me if it's wrong.
Session ID:- Whenever anyone first opens a website, that website sends a key to user end, and it is stored in cookie, and that key is used to pass with User ID and Password to Login. And this happens only once, that is, first time login, after that user can do other operations (or requests), but no such key is send with each such request. This key is session id.
Bearer Access Token:- Whenever anyone first opens a website, the user will then make a login, but no such key or token will pass with User Id and Password. After successful login, then that website will send a key, and this key is used (or passed) with each operation (or request) the user does after successful login. This key is bearer access token.
Now, please put light on CSRF and OAuth 2.0 in simpler way, because I am getting confused with what written on your blog for OAuth 2.0.
CSRF Token: Understand with example: https://stackoverflow.com/questions/5207160/what-is-a-csrf-token-what-is-its-importance-and-how-does-it-work
OAuth 2.0 is nothing but framework and technique to produce the security code. Just the HTTP and HTTPS define some rules (called protocols) for communication, similarly OAuth is used to define security type, level and approach.
The post "Access Token Issue" (https://perfmatrix.blogspot.com/2016/09/jmeter-oauth-20-login-issue.html) has not been written for explaining the type of security. It just provides the solution on how to handle authorization bearer token.
One question regarding JMeter Distributed Testing CSV Data Set:-
Suppose, I am doing a distributed testing on a website for 10 Users using JMeter, with 1 master and 1 slave, and in CSV Data Set, I have 10 UserName and Password. I am using 5 Users in thread group, as, master will put 5 users and slave will put 5 users, so total 10 will come. This is happening also. I am getting 10 users hit in Aggregate Report. But problem is:- when I am running the test, master machine is also using the first 5 UserName and Password, and Slave Machine is also using the Same first 5 USerName and Password from CSV. So, what happening here is, some good application like banking will only allow one username and password per session. Means, if you are logged in with your username and password, and someone else also try to login on same website in same time after you logged in using your username and password, he will get error in logging in. Same is happening in my case of distributed testing, both machine are using same first 5 username and password from CSV, and hence error is coming. So, any idea how to overcome this? Or, simplifying this, suppose I have 1 master and 2 slave and need to put 15 user load, 5 from each machine, then you to configure the CSV so that each machine will take 5 different records from CSV with 15 record so that login error does not comes up?
Request you to post your question on the relevant page only. If its JMeter question then I would suggest to post your query on JMeter page only; not on LoadRunner page. Such approach will help others to read your respective question and get the proper answer.
Okay Sir, I have got the answer for my last question regarding distributed testing with Jmeter. The answer is at below link:- https://stackoverflow.com/questions/2086216/distributed-jmeter-tests-with-per-slave-csv-datasets
In case if you have anyother approach, please share it, or else just publish my both request regarding this so that people come to know.
We are trying to record an Citrix application(hosted on web) using multi-protocol in Loadrunner. we are able to launch the application, login to Citrix and Open an app present in the Citrix (Eg: Calculator, One ESRI Application etc). Once after launching our App, we are not able to Enter Username/Password and in Calculator we are not able to enter any value/do any calculation. Everything seems to be disabled and not proceed further. Can you please suggest if you have any solution for this.
I did not get a deep level chance to work on Citrix protocol. The knowledge I gathered from a small assignment is documented in the below post: https://perfmatrix.blogspot.com/2017/04/loadrunner-citrix-protocol.html
Check on LoadRunner WhatsApp group, if you get help from anyone!
Hi, can anyone solve these scripting challenges and come up with a solution. It will be great if you write the function or code to solve the below scripting challenges:-
I have used lr_exit() in my script to exit an iteration. When ever this function is called the "Pacing" is discarded by the load runner while running a test in performance center. I mean to say if I set pacing of 10 seconds , it skips the pacing and start next iteration immediately. However , it is working fine in vugen. I have checked run time settings in PC as well.Has anybody faced this issue.
If you are using Web HTTP/HTML protocol then note down LR does not work on real browser, so there is no opening or closing the browser step is recorded. These are client side activities.
Excellent article!!! LoadRunner is popular automation testing tool used for validating a software application/system under load. It delivers most precise information about the performance, functionality and behavior of the software product. Loadrunner Training in Chennai | Loadrunner training Chennai|Qtp training in Chennai
ReplyDeleteHi Gagandeep,
ReplyDeleteCan you please share a blog on recording options of LR like url and html recording etc..
Hi Komal,
DeletePleas refer below URL:
https://perfmatrix.blogspot.in/2017/03/hp-loadrunner-script-recording-basic.html
Hi,
ReplyDeleteCan you please help me with below queries,
How to capture time download time of excel/txt file from SAP GUI?
I haven't face such scenario, so do not have practical experience. I googled and found that there is no such function in SAP GUI which we can use for file download time. Could you please raise a ticket to Microfocus may they have some solutions?
DeleteKindly share if you get any solution, this will help others.
Since Mercury days you can purchase a premium protocol for SAPGUI. It's about double the cost of a standard HTTP license. This is because SAP communicates via RPC rather than standard HTTP. It works like a charm.
ReplyDeleteit is nice article and eassy to understand there is any possible ways to download this as pdf.
ReplyDeleteif there is any possible please send the link
Thanks Bandarupalli!
ReplyDeleteNot right now, but I got some requests to document these posts and make them handy. I will try to prepare a handy document and publish it soon (when I get some free time).
Very Nice and easy to understand. If you can make some videos as well that would be awesome. Thumbs up for you. Cheers.
ReplyDeleteThanks Ashu!
DeleteSure, I will.
Hi Gagan,
ReplyDeleteA question to you.
i have a script ran with 100 users on controller by using 2 LGs,once test complete how i can get response times of my scripts based on LG wise? or where i can see those such type of results in analysis?
Hi Satish,
DeleteYou can not filter the result on the basis of LGs, because on completion of the test, reports from all the LGs are collected by Controller and prepared a combined report.
Thanks for the info Gagan, Is there any way to get before test or after test by doing any setup?
ReplyDeleteIf you are referring to your above comment then I do not think so there is such kind of set-up.
DeleteHi Gagan,
DeleteI have not worked in controller but in performance center one set up exists where we can filter based on hosts.I am not sure it will help or not.
Are you referring to the Analysis Tool filter based on LG?
DeleteDear Gagan,
ReplyDeleteThanks for the wonderful detailed article. Request you to please share an article on LR Truclient protocol.
Sure, I will.
DeleteHello Sir,
ReplyDeleteYour blog is wonderful.
I want to ask solution to a scenario.
Suppose, there is a scenario where a user searches flight between 2 cities. Now entering those details and clicking Search, he will get many response as multiple flights will be coming. Now he will select any one and book it.
Now the scenario is, suppose we are running the above script with 10 users parallely, they will enter cities details to find flight, now what I want is each user selects a unique flight to book, like as, 1st user books 3am flight, 2nd user books 5am flight, 3rd user books 8am flight, 4th user book 10am flight, and like that. So how to achieve this scenario. I.e. how to capture all response and pass them as parameter to each user differently. Hope this is clear.
Thanks.
I got your email on the same query, so replied the solution on email.
Deletecan you please give the solution for above scenario i too faced same problem could you please help me
DeleteThe easiest solution is to get the flight-related info in a parameter file and then use as per your requirement.
DeleteElse, the first user will pick the flight-related information using correlation and write them in a file which is shared between all the users. This file file be your parameter file with sequential order setting. When another user come and book the flight he will pick the next flight listed in the file.
Alternate Solution: If you your requirement fulfills then you can also capture the flight details using correlation and then pass randomly to the each user during runtime.
Thanks Sir.
DeleteI am posting the answer how this type of scenario can be achieved in JMeter. I have got the solution for this from Gagan Sir's blog itself. The link for the answer is:-
Deletehttps://perfmatrix.blogspot.com/2017/04/apache-jmeter-nested-variable-handling.html
Here, what we need to do is, we need to capture all the response values using Regular Expression in a variable and using -1 for matches. This will capture all such vales in that variable. Now, when we debug, we will get all those as variable_1, varibale_2, varibale_3, so on. Now we need to use the above said -V() function and parametrise this 1, 2, 3 according to the Users. The final parameterized function that need to be used where we are passing that response will be like ${__V(variable_${Values}), here ${Values} is for 1, 2, 3, etc passed form CSV file with respect with Users. Now, when we will run this, first user will select first flight, second user will select second flight, third user will select third flight and so on.
The above is for JMeter.
However, to achieve the same in Loadrunner (apart from above two solutions i.e. 1) parameterizing from file, and 2) passing randomly), I have asked Gagan Sir, hope he will come up with a similar solution for nested variable handling in Loadrunner.
Thanks.
Hi Kumar,
DeleteEvery tool has its own way of handling the common things. Like in JMeter dynamic value is handled by RegEx, but in LoadRunner it is primarily handled by correlation function. In LoadRunner, the above scenario can be handled by nesting the function; not the variable.
Example: lr_save_string(lr_paramarr_random("c_correlateParam"),"randParam");
To get more information on the same please refer: https://perfmatrix.blogspot.com/2017/04/random-value-from-correlation-parameter.html
Hi, is there any way to put wait timer in script till response is received.
ReplyDeleteHi Lakshmikant,
DeleteTimer (as a step) only executes when previous response is arrived completely. In case the request is taking more time to respond then you can increase step download timeout. But it does not have any connection with timer.
May be your case is like this:
Once an order is submitted then on the same page 3 states appeared.
1. Initialize
2. In Progress
3. Submitted
And you need to wait on the page until submitted state is appeared. In that case you need to write a logic to check the state on the page.
Let me know if your scenario is something else.
Hello Sir,
ReplyDeleteFor above scenario, can you share a snapshot or glimpse of that logic code which will check for that response, if you can?
Thanks.
Hi Kumar,
DeleteIt would be simply text check. Just you need to add web_reg_find() and provide the search for the state (like Submitted). Now you need to iterate this step until 'Submitted' word is found.
Thanks Lot..Very nice information on your blog on Performance testing tool for different tools. I felt something is missing like record settings. Please can you put that one with examples.
ReplyDeleteThanks Ramamurthy to highlight this. I posted one article HTML vs URL mode which comes under recording setting. https://perfmatrix.blogspot.com/2017/04/html-mode-vs-url-mode-recording-loadrunner.html
DeleteI will cover rest of the setting options in future post.
Hello Sir,
ReplyDeleteI have 2 queries to be asked.
1st.-> Suppose, we have to test an Application, let it be WebTours, for example. So, there, are basically 3 scenarios, first one is login->logout; 2nd one is login->browse flight->logout; and 3rd one is login->browse flight->purchase/book ticket -> logout.
Now, suppose while running for 1000 users to do load test, what will be the best approach to achieve the work load model from Controller.
Will it be like, record all scenarios in one script with each Scenario as Action, then in Controller, load the same script 3 times, just change the Action Scenario using Action_Block concept, then assign % or number of Users to each script, like 400 users for first Scenario, 300 users for second Scenario, and 300 Users for third Scenario, and then run all the 3 scripts at parallely. (Or we can record 3 scripts each for 3 scenarios and load them in controller and assign number or % of users and run them parallely).
Is this the best way to achieve the workload performance run? Please put light on this whether my approach is correct or not, if not correct, Please describe your approach?
Also, please let us know how to achieve the same in JMeter performance test?
Thanks.
Hi Kumar,
DeleteBoth the methods are correct, but I vote for first method as it is time saving and shows smart work.
Even in JMeter, the 3rd flow will be recorded once in which pages are separated by Transaction controller. Here we need use three thread groups and copy/paste the flow in all three thread groups. The transaction controllers can be disabled/delete where ever not required.
My 2nd Query is:-
ReplyDeleteSuppose we have to do distributed testing and we need to use load generators using JMeter & LoadRunner. So, my question is, to achieve distributed testing, is it necessary to place (or put) the VUser Script or JMX along with CSV data or Parametrization data in Same Place on Slave Load Generators, while doing distributed testing using both JMeter or LoadRunner, or is it like, it is required while doing from JMeter, but not from LoadRunner?
Please put a light on this.
Thanks.
In JMeter:
ReplyDelete1. Only data files like CSV Data set etc. need to place in the slave and same path needs to be given in the script. In case if you have more than 1 slave then the file must be placed at the same place in all the slaves.
2. No need to place script in the slave.
In LoadRunner:
1. LoadRunner transfers the Parameter files (.dat) to Load Generator before test starts, so no manual intervention is required.
2. If you have added the files (like image/pdf) in the script then again these files are transferred by LoadRunner itself to Load Generator.
3. While creating the script if you have given the absolute path of the file like (C:\LoadRunner\Project\myFile.txt) then you need to place the file in the load generator by creating the same folder structure.
4. No need to place the script at Load Generator, as it is LoadRunner's work.
Thanks Sir, it cleared a much general doubt.
ReplyDeleteHi Gagan sir,
ReplyDeletewhat is modular approach and what is use of it,can you please explain with one practical example.
Hi Harish,
DeleteCould you please clarify whether you are talking about custom module creation in C or you are referring to divide the script flow into different modules for re-use purpose?
Hi Sir,
ReplyDeleteI have one suggestion and few question to ask.
1.) Suggestion:- In Basics tab on this website, the PE Basic will come first and then PT Basic from top.
Now, questions:-
1.) Please put some light on both, one, custom module creation in C and how to use them w.r.t LR, and 2nd, dividing the script flow into different modules for re-use purpose, what is this?
2.) In your blog, how you managed to make dropdown for main tabs, any supporting code or setting for this?
3.) What is the major difference between Monitoring tool and Profiling tool with example tools name respectively?
4.) In worst situation, suppose we have get help from LoadRunner officials, so how to do that? In Jmeter, we need to send an email to user@jmeter.org, similarly, how to get help from LR Officials, or Official Community, what is the detailed procedure to write to them and get help?
Thanks.
Hi Kumar,
DeletePlease find my point-wise reply:
Suggestion:
1. May I know why you are suggesting to swap the position please?
Question:
1. It is as same as dividing the modules in c++ and java. Practically, I do not see anybody who use the modular approach in LoadRunner. Because script recording is easier than customization using modules.
Also, if you already have a script and want to use (say login module), then copy the login code and paste in the another script is a good option. This will also make the script fast rather than calling a module. (As per my view only)
2. May I know the reason for asking this question please?
3. Monitoring tool are platform independent and supported (nearly) all the technology. The main purpose of monitoring tool is check the performance of resources like CPU, Memory etc.
Example of Monitoring tool is: Micro Focus Site Scope
Profilling tool is specific to technology and helps to identify the performance by running some specific code to measure performance at code level.
Example of Profiling Tool: JProfiler
4. You must have SAID (support agreement ID) or Passport ID for that. SAID is provided by LoadRunner once you purchase the licenses.
Still, you can try to create an account for Micro Focus and check whether you can able to raise a service request without SAID.
Hi Kumar,
DeleteI got your reply. Please contact me on WhatsApp for further help.
As requested I am not publishing your last reply here.
Hi Sir,
ReplyDeleteOne more Question:-
Difference between Session ID, Bearer Access Token, CSRF Token and OAuth 2.0?
As per my work experience, I know 1st two, and writing it here, plz correct me if it's wrong.
Session ID:- Whenever anyone first opens a website, that website sends a key to user end, and it is stored in cookie, and that key is used to pass with User ID and Password to Login. And this happens only once, that is, first time login, after that user can do other operations (or requests), but no such key is send with each such request. This key is session id.
Bearer Access Token:- Whenever anyone first opens a website, the user will then make a login, but no such key or token will pass with User Id and Password. After successful login, then that website will send a key, and this key is used (or passed) with each operation (or request) the user does after successful login. This key is bearer access token.
Now, please put light on CSRF and OAuth 2.0 in simpler way, because I am getting confused with what written on your blog for OAuth 2.0.
Thanks.
CSRF Token: Understand with example: https://stackoverflow.com/questions/5207160/what-is-a-csrf-token-what-is-its-importance-and-how-does-it-work
DeleteOAuth 2.0 is nothing but framework and technique to produce the security code. Just the HTTP and HTTPS define some rules (called protocols) for communication, similarly OAuth is used to define security type, level and approach.
The post "Access Token Issue" (https://perfmatrix.blogspot.com/2016/09/jmeter-oauth-20-login-issue.html) has not been written for explaining the type of security. It just provides the solution on how to handle authorization bearer token.
DeleteSo Sir, technically, bearer access token authorization mechanism is an example of OAuth 2.0 framework, right?
DeleteYes, correct!
DeleteHi Sir,
ReplyDeleteOne question regarding JMeter Distributed Testing CSV Data Set:-
Suppose, I am doing a distributed testing on a website for 10 Users using JMeter, with 1 master and 1 slave, and in CSV Data Set, I have 10 UserName and Password. I am using 5 Users in thread group, as, master will put 5 users and slave will put 5 users, so total 10 will come. This is happening also. I am getting 10 users hit in Aggregate Report. But problem is:- when I am running the test, master machine is also using the first 5 UserName and Password, and Slave Machine is also using the Same first 5 USerName and Password from CSV. So, what happening here is, some good application like banking will only allow one username and password per session. Means, if you are logged in with your username and password, and someone else also try to login on same website in same time after you logged in using your username and password, he will get error in logging in. Same is happening in my case of distributed testing, both machine are using same first 5 username and password from CSV, and hence error is coming.
So, any idea how to overcome this?
Or, simplifying this, suppose I have 1 master and 2 slave and need to put 15 user load, 5 from each machine, then you to configure the CSV so that each machine will take 5 different records from CSV with 15 record so that login error does not comes up?
Thanks.
Hi Kumar,
DeleteRequest you to post your question on the relevant page only. If its JMeter question then I would suggest to post your query on JMeter page only; not on LoadRunner page. Such approach will help others to read your respective question and get the proper answer.
Thanks
Okay Sir, I have got the answer for my last question regarding distributed testing with Jmeter.
ReplyDeleteThe answer is at below link:-
https://stackoverflow.com/questions/2086216/distributed-jmeter-tests-with-per-slave-csv-datasets
In case if you have anyother approach, please share it, or else just publish my both request regarding this so that people come to know.
Thanks.
Hi Gagan,
ReplyDeleteWe are trying to record an Citrix application(hosted on web) using multi-protocol in Loadrunner. we are able to launch the application, login to Citrix and Open an app present in the Citrix (Eg: Calculator, One ESRI Application etc). Once after launching our App, we are not able to Enter Username/Password and in Calculator we are not able to enter any value/do any calculation. Everything seems to be disabled and not proceed further. Can you please suggest if you have any solution for this.
Thanks,
Hemanth P
I did not get a deep level chance to work on Citrix protocol. The knowledge I gathered from a small assignment is documented in the below post:
Deletehttps://perfmatrix.blogspot.com/2017/04/loadrunner-citrix-protocol.html
Check on LoadRunner WhatsApp group, if you get help from anyone!
Thank you Gagan :)
DeleteWhat is the difference between load testing and performance testing
ReplyDeleteHi Trinadh,
DeletePlease refer to the link:
https://perfmatrix.blogspot.com/2017/01/type-of-performance-test.html
Hi,
ReplyDeletecan anyone solve these scripting challenges and come up with a solution. It will be great if you write the function or code to solve the below scripting challenges:-
https://www.myloadtest.com/training/correlation-challenge/?step=1
https://www.myloadtest.com/training2/card-validation.php
https://www.myloadtest.com/scripting-exercise-a-basic-ajax-application/
https://www.myloadtest.com/scripting-exercise-unique-usernames/
Your blog is more informative, thanks for sharing valuable information about Load runner . Keep sharing...
ReplyDeleteThanks!
DeleteI have used lr_exit() in my script to exit an iteration. When ever this function is called the "Pacing" is discarded by the load runner while running a test in performance center.
ReplyDeleteI mean to say if I set pacing of 10 seconds , it skips the pacing and start next iteration immediately. However , it is working fine in vugen. I have checked run time settings in PC as well.Has anybody faced this issue.
Hi Sudhir,
DeletePlease use lr_exit(LR_EXIT_ITERATION_AND_CONTINUE, LR_AUTO);
Hi Sir, Need Help
ReplyDeleteIs there any code or way in LR which can close my browser after completing all actions.
If you are using Web HTTP/HTML protocol then note down LR does not work on real browser, so there is no opening or closing the browser step is recorded. These are client side activities.
DeleteThanks for sharing your valuable information and time.
ReplyDeleteLoadrunner Training in Delhi
You can approach to me directly if you want to spread the training details with wider audience. I will published your institute details.
Delete