Hello Raghav, This is really very informative video, thanks for all your hard effort to make such a wonderful video. I need your help and support to start 1 project, requirements are given below. Requirement : 1. Total number of business scenarios 17 2. Total number of users (threads) 60K (load) 3. Wants to distribute the load in each scenarios by percentage. 4. All the scenarios should be executed in parallel for the period of 1 hour in steady state. 5. Ramp up users until target load is reached and run full load for 60 minutes duration. 6. Test should be executed 1 cycle. 7. Application should be able to handle 60K virtual user with 381018 transaction per hour Queries: 1. Which Thread Group should I use? (Normal or concurrency thread group) 2. What are the Controller should I use to achieve the required output. 3. How to distribute the load in 19 Slave server. 4. How to pass the test data (60K users with id and password) 5. Currently its implemented with If single thread group using controller, but client wants to implement it with multi threading. I really appreciate your efforts and hard work , please help me to start this project. Thanks in advance.
Hi Arshad Here are some answers to your queries: 1. Which Thread Group should I use? (Normal or concurrency thread group) You should use the Concurrency Thread Group. The Concurrency Thread Group allows you to specify the number of users (threads) that will be running your test in parallel. This is the best option for your test because you want to distribute the load in each scenario by percentage. 2. What are the Controller should I use to achieve the required output? You should use the Loop Controller and the While Controller. The Loop Controller will allow you to run your test for a specified number of times. The While Controller will allow you to run your test until a specified condition is met. 3. How to distribute the load in 19 Slave server? You can use the Load Balancing Controller to distribute the load in 19 Slave servers. The Load Balancing Controller will distribute the load evenly across the slave servers. 4. How to pass the test data (60K users with id and password)? You can pass the test data (60K users with id and password) using the CSV Data Set Config element. The CSV Data Set Config element will allow you to read the test data from a CSV file. 5. Currently its implemented with If single thread group using controller, but client wants to implement it with multi threading. You can implement the test with multi threading by using the Parallel Controller. The Parallel Controller will allow you to run multiple thread groups in parallel. Here are some additional things to keep in mind: * The Concurrency Thread Group should be set to "Start concurrent threads". * The Loop Controller should be set to "Loop iterations". * The While Controller should be set to "While condition is true". * The Load Balancing Controller should be set to "Round robin". * The CSV Data Set Config element should be set to "Read CSV file". * The Parallel Controller should be set to "Start parallel threads". * I hope this helps
Hi Raghav, Thanks for uploading informative tutorials. It's a gr8 help for beginners. Can you please advise on REST API performance testing in real world scenario and the performance bench marks used for the mobile app performance using REST API's for Backend communication.
Nice informative videos. Thanks Raghav ............... I need to know what are the limitations of Jmeter as compared Load Runner or any Commercial performance tool...........
hi Mahendihasan, i am glad you found it useful. In my view i do not see a major limitation of JMeter over Load Runner or similar commercial tools. Of course with commercial tools we get some direct click and play features. However with JMeter having a great open source community and great plugins being developed continuously, i believe JMeter has far more advantages than limitations.
H Raghav, If we have 3 separate module scripts (Login, Create, & Manage) and trying to execute the same in 1 test plan using 3 different throughput controllers then we have to mention all the CSVs used under each module script in the same test plan.
Hi Raghav , Thanks for the amazing video. Can you please explain how can I achieve a specific throughput while doing a load test in Jmeter. On what all parameters is througput dependent like (no of threads , Ram duration etc ..?)
Hi Bhawna, In JMeter, you can achieve a specific throughput by adjusting the number of threads, ramp-up time, and duration of your test. Throughput is the number of requests per unit of time that a server can handle. It is dependent on the following parameters: Number of threads: The number of threads determines the number of concurrent users that JMeter simulates. Each thread represents a user, and JMeter sends requests to the server on behalf of each user. The more threads you use, the higher the throughput you can achieve. Ramp-up time: The ramp-up time is the time taken by JMeter to create all the threads. It determines how quickly the load increases. For example, if you set a ramp-up time of 10 seconds and the number of threads to 100, JMeter will create 10 threads per second. Test duration: The test duration determines the length of time for which JMeter will run the test. The longer the test duration, the more time JMeter has to send requests to the server, which can increase the throughput. Response time: The response time is the time taken by the server to respond to a request. If the response time is high, it can reduce the throughput. To achieve a specific throughput in JMeter, you can follow these steps: Determine the desired throughput: First, you need to determine the desired throughput. For example, if you want to achieve a throughput of 500 requests per minute, you need to send 8.33 requests per second. Set the number of threads: Based on the desired throughput, set the number of threads that JMeter needs to simulate. You can use the formula: (Desired Throughput / (Response Time * Number of Threads)) = Requests per Second. Set ramp-up time: Set the ramp-up time based on the number of threads you have set. If you have set 100 threads, and you want to ramp-up in 10 seconds, then set the ramp-up time to 10 seconds. Set test duration: Set the test duration to a time period that is long enough to collect meaningful results. By adjusting these parameters, you can achieve the desired throughput in JMeter
Hi Raghav i like your most of the video i watched so far . all the videos are very useful and full of information. when you say Distributed Load Test i misunderstand . is there possible to Distribute the load test in multiple desktop . would be great if you will add any video please. many thanks .
Hi raghav , your all videos are really understandable n helpful for everyone ,thank u for posting these videos. Raghav , i want to perform distributed testing(on e commerce site) using master and slaves configuration on my single m/c ,is it possible. pls help me .
Hi Raghav, thanks for this tutorial. This method works unexpectacly if I'll add loop count or Infinite+duration. Please kindly explain to me how to config test plan with dividing of thread group with configured loop count or Infinite+duration. Thanks in advance!
Hi Raghav, Thanks for the video and it worked for me. I have used CSV data set config with more number of different users.And i have crossed checked the Login &logout details on DB (On Audit log) the mentioned users are there. Here my question is, i need to give load for the particular page and how could i know that capacity of the particular page & and where i have to view the exact results for that page.
Hi Hanif, you can setup your page in request sampler. For capacity, you have to test with different loads. Keep increasing the load gradually and note the time on each iteration. When you see some major deviation, you can know the capacity of the servers.
Hi Raghav, firstly thanks for sharing all informative channels. I have one of web application performance requirement is below 1. Login page has configured with IDP OAuth - [using jmeter IDP auth worked] 2. Load Page A, Page B and Page C - verify each page loading response.( front end application designed with nodejs POLYMER customized components and in DOM loads with N no.of #shadow-root elements) to click any icon or button and find element for that DOM loads with levels of shadow-root are minimum 10 to 20 [ using jmeter able to traverse each page after login] 3. Now, I want to update some attribute information. How to achieve this ? (tried blaze meter recording, but recording not working after login page due to may be shadow-root elements). Suggest me how to proceed for perf testing when application html5 DOM enabled with shadow-root ? if you have any such videos please share with us.
thanks for your message Pradeep, You can see here: JMETER BEGINNER th-cam.com/play/PLhW3qG5bs-L-zox1h3eIL7CZh5zJmci4c.html JMETER INTERMEDIATE th-cam.com/play/PLhW3qG5bs-L-PpK83CC_m_bWd_722tvgk.html JMETER ADVANCED th-cam.com/play/PLhW3qG5bs-L_Eosy1Nj1tKHC5jcBAVkPb.html JMETER TIPS & TRICKS th-cam.com/play/PLhW3qG5bs-L-oMT1Cnrurt7gEaLrRub2V.html
Hi Sir, Very nice to see your videos. I have a question i want to test my spark and flink custom applications for throughput and latency testing in cluster mode. How i can use jmeter for this purpose? Please give an idea . Need guidline
When using JMeter for load testing, you can achieve sequential execution of samplers even with distributed user load. Let's break down the steps: 1. Default Sequential Execution: By default, JMeter samplers run in sequential order within a thread group. However, when you add multiple users (threads) to the test run, each user might be at different steps simultaneously. This can create the appearance of random execution order (e.g., 1-3-2-4) due to parallelism¹. 2. Synchronizing Timer: To ensure that all users proceed sequentially through your samplers, you can use the Synchronizing Timer element. Here's how: - Add a Synchronizing Timer under each HTTP Request sampler. - Set the Number of Simultaneous Users to Group by to the desired value (e.g., the total number of threads in your thread group). - If you set it to 0, it's equivalent to the total number of threads in the thread group¹. 3. Example Scenario: Let's say you have three thread groups: - Thread Group 1: Execute once. - Thread Group 2: Simultaneously execute with 500 users for load testing. - Thread Group 3: Execute once. The threads within these groups are dependent on each other, ensuring sequential execution². Remember that JMeter provides various elements to control execution order, and the Synchronizing Timer is a powerful tool for achieving sequential behavior in a distributed load testing scenario ..
Hi All the videos are very informative. I have one question for you. Can you please help me how to run more than one HTTP Request(Sampler) at the same time or say parallelly? or can direct me in which video you have covered this.
Hi Anwar, check if this helps - stackoverflow.com/questions/36188913/running-all-http-request-at-one-time-in-jmeter stackoverflow.com/questions/28938841/how-to-send-concurrent-requests-in-jmeter
Thanks for detailed video. Can you please explain how to handle below scenario using jmeter. I need to do load testing on Job portal. scenario is 200 user for registration, 200 users for job searching, 200 users for Login and 200 users for Job posting and each 200 user need to be do their actions (registration, searching,login and job posting). please explain how to create test plan for this scenario ,Thanks in Advance
Hi Raghav, thanks for sharing the useful videos. Need some help for the actual distributed testing with JMeter like the Master having more Slaves(1 or more) and targetting 1 server. What configurations we have to do with which files.
Hi Deepak, it depends on the requirements. Of course all the features will not be needed for a single test so you have to see what you need and prepare the test accordingly.
Hi raghav , thanks for the video but I have a query if we want to have different users for all the scenarios..what condition we should keep? because in this it may happen that the user it took for the first thread group , the same user is taken for the second or third thread group.
hi raghav, I have also one question. My scenario-login,registration(form is to long. so it distributed in three pages.like page1,page2,page3. So firstly user fill entries on page 1 then press continue button after that he will reach on page 2,same as on page 2 and page3),report1,report2,logout. I want to test this scenario for 100users in which all 100 users login at same time,then from 100 users 60 users do registration in sequential manner,20 user see report 1 and rest user see report.then all 100 users logout at same time
Hi Raghav , do we have controller or anything through which i can check my server max throughput . I used constant throughput but with this as well I couldnt achieve my desire hits value (My devops is saying 100hits/sec application can survive).I want to assure this.And also let me know the correct place where i can apply it.
Hi Raghav, Videos are very informative and appreciate the work and time you had put in. Can you add a video on the jmeter and Git integration?Can we push the .jmx files and results.jtl files to repo-github and then execute the jenkins job by pulling the files from the github. Is that possible?
Hi Raghav,nice explanation. Threadgroup-10 users throughput controller 1- 60% throughput controller 2- 40% I have 3 samplers/requests in throughput controller1 and 2 samplers/requests in throughput controller2.And also i have correlated some values in requests of throughput cont1 and called them in the requests of throughput cont2.But the correlated values in throughput cont1 are not passing to the requests in throughput cont2,so that all the requests in throughput cont2 are failing. And also it is working fine if i give same percentage to throughput cont 1 & 2.The problem comes when i give different percentages. Please suggest a solution for this. I have csv data set config also in my script.
Hi Raghav, in one test plan I will merge different scripts with different csv files, is this right procedure..I have a save script, submit script and update script, each files have different csv files, I will add a master thresgropup, three different throughput controller with percentage distribution ans under each throughput controller csv configuration will come..pls reply with your insights
H Sudha, Using different or same csv file will be as per your requirement. Also the controllers should be configured as per your application and performance test requirements. A basic rule in performance testing is to have a configuration that can be as close as real world scenario. Also you can configure listeners as per your report requirements. suggestion : disable any memory consuming listeners like view results tree, graphs, etc..
Hi Raghav! Your video is nice and informative, however, the title is a bit misleading. You're referring to (and explaining) "Request Distribution", "User Profile" and, according to the title, most people will come here thinking they'll see the setup for JMeter client / server topology with multiple Load Generators (slaves) connected to the same Controller (master). As per (1 example): www.blazemeter.com/blog/how-to-perform-distributed-testing-in-jmeter Comming back to the subject, an interesting detail, related to Throughput Controller in JMeter is the fact that given under a node, let's say thread group, such as bellow, you might see some different behavior between HPLR and JMeter and this small piece of info, in return, might save some money, time and stress for perf engineers. JMeter Thread Group >> Throughput Controller 1 - 30% >>>> request_A >> Throughput Controller 2 - 20% >>>> request_B >> Throughput Controller 3 - 50% >>>> request_C In the above piece of pseudo-code, in JMeter, there is a possibility that a VU will do, at some point, on some iteration one of the below combinations: - request_A, request_B, request_C - request_A, request_B - request_A, request_C - request_B, request_C - request_A - request_B - request_C - none (in theory) After 10 iterations with 1 VU: You'll have at least 10 requests Note: If checked, "Per User" will cause the controller to calculate whether it should execute on a per user (per thread) basis. If unchecked, then the calculation will be global for all users. For example, if using total execution mode, and uncheck "per user", then the number given for throughput will be the total number of executions made. If "per user" is checked, then the total number of executions would be the number of users times the number given for throughput. HPLR Group Properties (Run logic Random) >> Action 1 - Random percents 30% >>>> request_A >> Action 2 - Random percents 20% >>>> request_B >> Action 3 - Random percents 50% >>>> request_C in HPLR, when editing the Runtime settings > Run logic, if you add a Group Properties block and establish Run Logic >, modeling the same 30%, 20%, 50% Request Distribution, one Virtual User can do just the below combinations: - request_A - request_B - request_C After 10 iterations with 1 VU: You'll have exactly 10 requests The different behavior is quite important and, at times, you'll have to work a bit on readapting the scripts when moving from HPLR to JMeter and the other way around. Cheers!
![POLYCAT - ข้อความรูปยิ้ม [Music Video]](http://i.ytimg.com/vi/lBhIgPGCgbI/mqdefault.jpg)
I`am finished beginner and now intermediate. It was very helpfull and thx for your hard work :)
Glad it was helpful, thanks
Thank you for the detailed explanation
You are welcome
That was very informative! Thank you very much!
You're welcome Paulo
Very nice explanation, easy to understand, Thanks Raghav...
You're welcome Yakini
Very informative video series
Hello Raghav, This is really very informative video, thanks for all your hard effort to make such a wonderful video. I need your help and support to start 1 project, requirements are given below.
Requirement :
1. Total number of business scenarios 17
2. Total number of users (threads) 60K (load)
3. Wants to distribute the load in each scenarios by percentage.
4. All the scenarios should be executed in parallel for the period of 1 hour in steady state.
5. Ramp up users until target load is reached and run full load for 60 minutes duration.
6. Test should be executed 1 cycle.
7. Application should be able to handle 60K virtual user with 381018 transaction per hour
Queries:
1. Which Thread Group should I use? (Normal or concurrency thread group)
2. What are the Controller should I use to achieve the required output.
3. How to distribute the load in 19 Slave server.
4. How to pass the test data (60K users with id and password)
5. Currently its implemented with If single thread group using controller, but client wants to implement it with multi threading.
I really appreciate your efforts and hard work , please help me to start this project. Thanks in advance.
Hi Arshad
Here are some answers to your queries:
1. Which Thread Group should I use? (Normal or concurrency thread group)
You should use the Concurrency Thread Group. The Concurrency Thread Group allows you to specify the number of users (threads) that will be running your test in parallel. This is the best option for your test because you want to distribute the load in each scenario by percentage.
2. What are the Controller should I use to achieve the required output?
You should use the Loop Controller and the While Controller. The Loop Controller will allow you to run your test for a specified number of times. The While Controller will allow you to run your test until a specified condition is met.
3. How to distribute the load in 19 Slave server?
You can use the Load Balancing Controller to distribute the load in 19 Slave servers. The Load Balancing Controller will distribute the load evenly across the slave servers.
4. How to pass the test data (60K users with id and password)?
You can pass the test data (60K users with id and password) using the CSV Data Set Config element. The CSV Data Set Config element will allow you to read the test data from a CSV file.
5. Currently its implemented with If single thread group using controller, but client wants to implement it with multi threading.
You can implement the test with multi threading by using the Parallel Controller. The Parallel Controller will allow you to run multiple thread groups in parallel.
Here are some additional things to keep in mind:
* The Concurrency Thread Group should be set to "Start concurrent threads".
* The Loop Controller should be set to "Loop iterations".
* The While Controller should be set to "While condition is true".
* The Load Balancing Controller should be set to "Round robin".
* The CSV Data Set Config element should be set to "Read CSV file".
* The Parallel Controller should be set to "Start parallel threads".
*
I hope this helps
Hi Raghav, Thanks for uploading informative tutorials. It's a gr8 help for beginners. Can you please advise on REST API performance testing in real world scenario and the performance bench marks used for the mobile app performance using REST API's for Backend communication.
Hi Vaishali, I will cover these concepts in a session
very informative thank you for your effort
Glad it was helpful Ritesh
Nice, useful and clear, many thanks!
You're welcome Ola
Well Explained..
Thanks Jagan
nice one.. thanks buddy
You're welcome
Nice Explanation, thanks buddy
You're welcome Kumar
thanks raghav
Nice informative videos. Thanks Raghav
...............
I need to know what are the limitations of Jmeter as compared Load Runner or any Commercial performance tool...........
hi Mahendihasan, i am glad you found it useful.
In my view i do not see a major limitation of JMeter over Load Runner or similar commercial tools. Of course with commercial tools we get some direct click and play features. However with JMeter having a great open source community and great plugins being developed continuously, i believe JMeter has far more advantages than limitations.
Hi sir,
All the videos are very very helpful.So, Hates off for you. Sir will you please also explane the PRE and POST Processor .
Thanks
Thanks Satender for watching, Will surely do some videos on that.
Thanks raghav
H Raghav, If we have 3 separate module scripts (Login, Create, & Manage) and trying to execute the same in 1 test plan using 3 different throughput controllers then we have to mention all the CSVs used under each module script in the same test plan.
Hi Archana, yes
More Informative bro thanks a lot..
Thanks for watching.
Hi Raghav , Thanks for the amazing video. Can you please explain how can I achieve a specific throughput while doing a load test in Jmeter. On what all parameters is througput dependent like (no of threads , Ram duration etc ..?)
Hi Bhawna,
In JMeter, you can achieve a specific throughput by adjusting the number of threads, ramp-up time, and duration of your test. Throughput is the number of requests per unit of time that a server can handle. It is dependent on the following parameters:
Number of threads: The number of threads determines the number of concurrent users that JMeter simulates. Each thread represents a user, and JMeter sends requests to the server on behalf of each user. The more threads you use, the higher the throughput you can achieve.
Ramp-up time: The ramp-up time is the time taken by JMeter to create all the threads. It determines how quickly the load increases. For example, if you set a ramp-up time of 10 seconds and the number of threads to 100, JMeter will create 10 threads per second.
Test duration: The test duration determines the length of time for which JMeter will run the test. The longer the test duration, the more time JMeter has to send requests to the server, which can increase the throughput.
Response time: The response time is the time taken by the server to respond to a request. If the response time is high, it can reduce the throughput.
To achieve a specific throughput in JMeter, you can follow these steps:
Determine the desired throughput: First, you need to determine the desired throughput. For example, if you want to achieve a throughput of 500 requests per minute, you need to send 8.33 requests per second.
Set the number of threads: Based on the desired throughput, set the number of threads that JMeter needs to simulate. You can use the formula: (Desired Throughput / (Response Time * Number of Threads)) = Requests per Second.
Set ramp-up time: Set the ramp-up time based on the number of threads you have set. If you have set 100 threads, and you want to ramp-up in 10 seconds, then set the ramp-up time to 10 seconds.
Set test duration: Set the test duration to a time period that is long enough to collect meaningful results.
By adjusting these parameters, you can achieve the desired throughput in JMeter
@@RaghavPal Thank you so much !
Please upload a video regarding pre-post processor( bean shell) in jmeter.
I will plan Monalika
@@RaghavPal Thanku so much!!..I hv watced all ur jmeter video..all are very good..and helped me alot to learn and implement in my project.
Hi Raghav i like your most of the video i watched so far . all the videos are very useful and full of information.
when you say Distributed Load Test i misunderstand . is there possible to Distribute the load test in multiple desktop . would be great if you will add any video please. many thanks .
Hi Vimal, we use master-slave setup for that. Will try to do a session on that.
Hi Raghav thank you so much for reply , Appreciated !! looking forward for your new vedio for Distributed Load Test on machine. have great day.
Thanks, Raghav
You're welcome Monika
Hi raghav , your all videos are really understandable n helpful for everyone ,thank u for posting these videos.
Raghav , i want to perform distributed testing(on e commerce site) using master and slaves configuration on my single m/c ,is it possible. pls help me .
Hi Swati, you can do distributed testing in JMeter, However it is done with multiple systems, where one is master and others are slaves.
yes, i read like that, but i just wanted to know , can i do the same on single m/c.
don't think so as it will have the same ip
I have not tried that
Hi Raghav, thanks for this tutorial. This method works unexpectacly if I'll add loop count or Infinite+duration. Please kindly explain to me how to config test plan with dividing of thread group with configured loop count or Infinite+duration. Thanks in advance!
Hi, I will check and plan a session on this
Hi Raghav,
Thanks for the video and it worked for me. I have used CSV data set config with more number of different users.And i have crossed checked the Login &logout details on DB (On Audit log) the mentioned users are there.
Here my question is, i need to give load for the particular page and how could i know that capacity of the particular
page & and where i have to view the exact results for that page.
Hi Hanif, you can setup your page in request sampler. For capacity, you have to test with different loads. Keep increasing the load gradually and note the time on each iteration. When you see some major deviation, you can know the capacity of the servers.
@@RaghavPalThanks Raghav
Hi Raghav, firstly thanks for sharing all informative channels.
I have one of web application performance requirement is below
1. Login page has configured with IDP OAuth - [using jmeter IDP auth worked]
2. Load Page A, Page B and Page C - verify each page loading response.( front end application designed with nodejs POLYMER customized components and in DOM loads with N no.of #shadow-root elements) to click any icon or button and find element for that DOM loads with levels of shadow-root are minimum 10 to 20 [ using jmeter able to traverse each page after login]
3. Now, I want to update some attribute information. How to achieve this ? (tried blaze meter recording, but recording not working after login page due to may be shadow-root elements).
Suggest me how to proceed for perf testing when application html5 DOM enabled with shadow-root ? if you have any such videos please share with us.
Hi Triveni, I do not have a 1st hand experience on this and will suggest to check online.
Thanks Raghav , you are doing a great job .
Also please tell me is there any advance tutorial session on jmeter
Thanks Once Again!
thanks for your message Pradeep, You can see here:
JMETER BEGINNER
th-cam.com/play/PLhW3qG5bs-L-zox1h3eIL7CZh5zJmci4c.html
JMETER INTERMEDIATE
th-cam.com/play/PLhW3qG5bs-L-PpK83CC_m_bWd_722tvgk.html
JMETER ADVANCED
th-cam.com/play/PLhW3qG5bs-L_Eosy1Nj1tKHC5jcBAVkPb.html
JMETER TIPS & TRICKS
th-cam.com/play/PLhW3qG5bs-L-oMT1Cnrurt7gEaLrRub2V.html
Thanks for the links raghav
Hi Sir, Very nice to see your videos. I have a question i want to test my spark and flink custom applications for throughput and latency testing in cluster mode. How i can use jmeter for this purpose? Please give an idea . Need guidline
Hi Aamer i have not done this. May be this can help - github.com/SparklineData/spark-druid-olap/wiki/Simple-Jmeter-Guide-in-Spark-Testing
Hello
Could you please help to understand how to make sequential execution of Sampler with distributed user load.
When using JMeter for load testing, you can achieve sequential execution of samplers even with distributed user load. Let's break down the steps:
1. Default Sequential Execution:
By default, JMeter samplers run in sequential order within a thread group. However, when you add multiple users (threads) to the test run, each user might be at different steps simultaneously. This can create the appearance of random execution order (e.g., 1-3-2-4) due to parallelism¹.
2. Synchronizing Timer:
To ensure that all users proceed sequentially through your samplers, you can use the Synchronizing Timer element. Here's how:
- Add a Synchronizing Timer under each HTTP Request sampler.
- Set the Number of Simultaneous Users to Group by to the desired value (e.g., the total number of threads in your thread group).
- If you set it to 0, it's equivalent to the total number of threads in the thread group¹.
3. Example Scenario:
Let's say you have three thread groups:
- Thread Group 1: Execute once.
- Thread Group 2: Simultaneously execute with 500 users for load testing.
- Thread Group 3: Execute once.
The threads within these groups are dependent on each other, ensuring sequential execution².
Remember that JMeter provides various elements to control execution order, and the Synchronizing Timer is a powerful tool for achieving sequential behavior in a distributed load testing scenario
..
@@RaghavPal Thank you so much for the detailed information.
Hi
All the videos are very informative. I have one question for you. Can you please help me how to run more than one HTTP Request(Sampler) at the same time or say parallelly? or can direct me in which video you have covered this.
Hi Anwar, check if this helps - stackoverflow.com/questions/36188913/running-all-http-request-at-one-time-in-jmeter
stackoverflow.com/questions/28938841/how-to-send-concurrent-requests-in-jmeter
Thanks for detailed video. Can you please explain how to handle below scenario using jmeter.
I need to do load testing on Job portal. scenario is 200 user for registration, 200 users for job searching, 200 users for Login and 200 users for Job posting and each 200 user need to be do their actions (registration, searching,login and job posting).
please explain how to create test plan for this scenario ,Thanks in Advance
Hi, you can add scripts in different samplers in a thread group and run it as per the load required
Hi Raghav, thanks for sharing the useful videos. Need some help for the actual distributed testing with JMeter like the Master having more Slaves(1 or more) and targetting 1 server. What configurations we have to do with which files.
Hi Mrutyunjay, i will try to publish a video soon on that
Automation Step by Step - Raghav Pal Thank you so much..
hey raghav canyou pllease elaborate ..........as to how do we handle concurrent ajax and javascript user.
hi Shivansh, can you please elaborate the requirement with some example. I will try to check it.
Do you a Test with all the explained features of Jmeter, that means it should be a Test suite ready to run.
Hi Deepak, it depends on the requirements. Of course all the features will not be needed for a single test so you have to see what you need and prepare the test accordingly.
Hi raghav , thanks for the video but I have a query if we want to have different users for all the scenarios..what condition we should keep? because in this it may happen that the user it took for the first thread group , the same user is taken for the second or third thread group.
Hi Anmol, you can use csv file for different values.
Automation Step by Step - Raghav Pal ok i will implement this.!! Thanks
hi raghav, I have also one question.
My scenario-login,registration(form is to long. so it distributed in three pages.like page1,page2,page3. So firstly user fill entries on page 1 then press continue button after that he will reach on page 2,same as on page 2 and page3),report1,report2,logout.
I want to test this scenario for 100users in which all 100 users login at same time,then from 100 users 60 users do registration in sequential manner,20 user see report 1 and rest user see report.then all 100 users logout at same time
Hi Sumit, you can create different thread groups and run them in parallel or use Module controller th-cam.com/video/jKiZ1efpE5w/w-d-xo.html
You are the best lol..
Thanks Sharath
Hi Raghav , do we have controller or anything through which i can check my server max throughput . I used constant throughput but with this as well I couldnt achieve my desire hits value (My devops is saying 100hits/sec application can survive).I want to assure this.And also let me know the correct place where i can apply it.
Hi Abhishek, can try throughput shaper timer - jmeter-plugins.org/wiki/ThroughputShapingTimer/
Hi Raghav, can you try this scenario (20%,30%, 50%) with a total of 11 users then 1 thread will be missing
will plan a session
Hi,
Will this work the same way when using a stepping thread group?
Hi Chris, Stepping Thread Group has a different use. Although you can use both of them to create a realistic performance test.
Hi Raghav
Does throughput controller pick up unique users for the distribution??
Deepak, will need to check to be sure.
Hi Raghav, Videos are very informative and appreciate the work and time you had put in. Can you add a video on the jmeter and Git integration?Can we push the .jmx files and results.jtl files to repo-github and then execute the jenkins job by pulling the files from the github. Is that possible?
Hi Bharat, yes you can. Infact the ideal solution will be to have Jenkins pull the latest file and then run the test.
Thanks Raghav...I dont see video for the same...appreciate if you can prepare a video which would be very helpful for all of us.
Hi Raghav,nice explanation.
Threadgroup-10 users
throughput controller 1- 60%
throughput controller 2- 40%
I have 3 samplers/requests in throughput controller1 and 2 samplers/requests in throughput controller2.And also i have correlated some values in requests of throughput cont1 and called them in the requests of throughput cont2.But the correlated values in throughput cont1 are not passing to the requests in throughput cont2,so that all the requests in throughput cont2 are failing.
And also it is working fine if i give same percentage to throughput cont 1 & 2.The problem comes when i give different percentages.
Please suggest a solution for this.
I have csv data set config also in my script.
Hi Manasa, pls do a debugging and check the logs. Use this session -
th-cam.com/video/S8ehjk9XWtE/w-d-xo.html
th-cam.com/video/7JI3GD0ifTw/w-d-xo.html
Hi Raghav, in one test plan I will merge different scripts with different csv files, is this right procedure..I have a save script, submit script and update script, each files have different csv files, I will add a master thresgropup, three different throughput controller with percentage distribution ans under each throughput controller csv configuration will come..pls reply with your insights
H Sudha, Using different or same csv file will be as per your requirement. Also the controllers should be configured as per your application and performance test requirements. A basic rule in performance testing is to have a configuration that can be as close as real world scenario. Also you can configure listeners as per your report requirements.
suggestion : disable any memory consuming listeners like view results tree, graphs, etc..
Hi Raghav!
Your video is nice and informative, however, the title is a bit misleading. You're referring to (and explaining) "Request Distribution", "User Profile" and, according to the title, most people will come here thinking they'll see the setup for JMeter client / server topology with multiple Load Generators (slaves) connected to the same Controller (master).
As per (1 example): www.blazemeter.com/blog/how-to-perform-distributed-testing-in-jmeter
Comming back to the subject, an interesting detail, related to Throughput Controller in JMeter is the fact that given under a node, let's say thread group, such as bellow, you might see some different behavior between HPLR and JMeter and this small piece of info, in return, might save some money, time and stress for perf engineers.
JMeter Thread Group
>> Throughput Controller 1 - 30%
>>>> request_A
>> Throughput Controller 2 - 20%
>>>> request_B
>> Throughput Controller 3 - 50%
>>>> request_C
In the above piece of pseudo-code, in JMeter, there is a possibility that a VU will do, at some point, on some iteration one of the below combinations:
- request_A, request_B, request_C
- request_A, request_B
- request_A, request_C
- request_B, request_C
- request_A
- request_B
- request_C
- none (in theory)
After 10 iterations with 1 VU: You'll have at least 10 requests
Note: If checked, "Per User" will cause the controller to calculate whether it should execute on a per user (per thread) basis. If unchecked, then the calculation will be global for all users. For example, if using total execution mode, and uncheck "per user", then the number given for throughput will be the total number of executions made. If "per user" is checked, then the total number of executions would be the number of users times the number given for throughput.
HPLR Group Properties (Run logic Random)
>> Action 1 - Random percents 30%
>>>> request_A
>> Action 2 - Random percents 20%
>>>> request_B
>> Action 3 - Random percents 50%
>>>> request_C
in HPLR, when editing the Runtime settings > Run logic, if you add a Group Properties block and establish Run Logic >, modeling the same 30%, 20%, 50% Request Distribution, one Virtual User can do just the below combinations:
- request_A
- request_B
- request_C
After 10 iterations with 1 VU: You'll have exactly 10 requests
The different behavior is quite important and, at times, you'll have to work a bit on readapting the scripts when moving from HPLR to JMeter and the other way around.
Cheers!
Sir also have some queries related to plugins. So,please give me your email id from where i will asked the questions.
Thanks Raghav
Satender, you can find the email in the about section of this channel.