My apologies if i may be reposting a problem which may already have been
addressed here.
But i was not not able to find a proper answer to my problem.
I am setting up JMETER in distributed mode across 10 different systems,
using OS process handler in a thread group.
The purpose is to invoke a executable and generate load for a server.
around 200 parallel executions are required on each system for a time frame
with ramp and down.
I have all this figured out thanks to quite an elaborate support forum and
documentation.
But out of these 10 systems, 1 specific system needs to run 50 parallel
executables and with some different parameters.
How can i select/force one specific system to do a different operation that
other 9 systems?
Please guide.
Regards
Sachin
If you know the machine name / ip in which you want to change the load, you can use User Defined Variables.
MachineA, MachineB.. are the names of the slave machines in which you are planning to run your jmeter test.
Then, you can get machine specific details by using
${__evalVar(${this.machine.users})}
If you use the expression in the No of threads in the ThreadGroup as given below, depends on the slave , for the same test, it puts different loads.
This is just an example. But similar approach should solve your issue.
Related
Running a JMeter distributed test on 10 machines, every machine with 25 users. While checking the report generated in the Time Vs Threads graph in the Response Times section it shows number of active threads 25. Actually it is showing 25 threads from every machine. I would expect it show max 250 threads and aggregate graph instead of plotting 10 points.
Can someone help how we can get aggregate graph for all threads.
Thanks in advance.
You are generating report from the system it is running. You need to consolidate all the reports from the system and then generate the graph using merge results
Now, if your setup consist of one master and other machines as slaves then slaves machines should send the listener data to master.
Check below link:-
https://jmeter.apache.org/usermanual/remote-test.html#sendermode
Try with master and slave for two machines, initially, and do consider the below info:-
Make sure that all the nodes (client and servers) :
are running exactly the same version of JMeter. are using the same
version of Java on all systems. Using different versions of Java may
work but is discouraged. have a valid keystore for RMI over SSL, or
you have disabled the use of SSL.
Hope this helps.
Try adding __machineName() or __machineIP() function so it would be possible to distinguish the results coming from different machines and properly calculate the concurrency.
More information:
Active Threads Over Time Listener
Apache JMeter Functions - An Introduction
This is for e-commerce project where the number of users login will be more. I have been given a benchmark 8000 concurrent users need to login and response time should be 3 minutes
#abi , hi .
Let me provide couple notes here.
Depending upon Your connection bandwidth , from my experience as performance test engineer, I'd say jmeter single instance usually holds up to 1k(1000)- 2k(2000) in best case users load.
Considering You have a requirement for 8k (8000 users) load, You need to launch jmeter in distributed mode ( master <-> slaves).
For this config setup I'd recommend to go with 1 master node and 4slaves. For that - You will need 5 machines (aws/azure, whatever) in the same sub-network.
Re more technical details on distributed setup, please take a look:
in public jmeter documentation
please also look into this step-by-step setup manual
Also, when i've been doing set-up for 10k load for one of my recent projects - I did couple notes for myself in g-doc . Let me know if it opens fine for You.
Last note, If You need to do some load/performance tests on APIs that require AUTHZ, I'd recommend to split authorize (IDP bypass) and performance scenario itself - in different thread groups. As usually IDP in DEVs/Stagings does not hold much load .
So at first You need to authorize w/o any load (1st Thread group).
And in 2nd Thread group - start calling target APIs under the test.
It depends on:
Your machine specifications (CPU, RAM, NIC card, hard drive, etc.)
The nature of your Test Plan (number of requests, size of requests/responses, number of pre/post processors, assertions, timers, etc.)
Response time of your application
So if your test is a simple GET request which returns small text response - you might simulate 10 000 of users on a mid-range modern laptop. And if your test is connected with heavy requests, large responses, file uploads, etc. - it might be 1000 users.
Make sure to follow recommendations from JMeter Best Practices
Make sure to have monitoring of resources usage of your system (CPU, RAM, Swap, etc.). You can use JMeter PerfMon Plugin for this.
Make sure that your test behaves like a real browser
Start with 1 virtual user and gradually increase the load until you reach 8000 virtual users or JMeter starts lacking resources, whatever comes the first. If you can simulate 8000 users from a single machine - you're good to go. If not - you will have to consider Distributed Testing.
What will be the hardware configuration required to run concurrent of 10,000 user's load in jmeter through non-gui mode?
There is no exact answer for this, but in my experience 10,000 users on a single instance doing anything other than very basic work will be too many.
You should look into setting up distributed testing, so that you have many different injectors. Without knowing anything about your application, I would still want at least 10 instances.
This link should get you started: http://jmeter.apache.org/usermanual/jmeter_distributed_testing_step_by_step.pdf
If you have budget for it, any of the cloud based Jmeter services will make it a lot less painless. Blazemeter is one such offering in this area: http://blazemeter.com/
I have created a test plan for creating userprofile.
I want to run my test plan for 100 users but when i run it for 10 users then it is running successfully with rump up time of 2 sec; but when i try it for 100 users & more than that it is getting failed I am giving rump uptime of 40 sec for 100 users.
I am not able to understand what may be the problem with it.
In my test plan the thread user are differentiated with id
Thanks in Advance.
It's a wide question, this behavior can be caused by
Your application under test can't handle load of 100 threads. Check logs for errors and make sure that application/web server and/or database configuration allow 100+ concurrent connections. Also you can check "Latency" metric to see if there is a problem with infrastructure or application itself.
Your load generator machine can't create 100 concurrent threads. If so - you'll need to consider JMeter Distributed Testing
Your script isn't optimized. I.e. using memory-consuming listeners like "View Results Tree", any graph listeners, regular expression extractors. Try following JMeter Performance and Tuning Tips guide and see whether it resolves your issue.
Agree with Dmitri, reason could be one of the above three.
One more thing you can try.
You can run your jmeter in ui mode for validation of your script and after validation you can run it in non-ui mode which will save lot of memory and cpu processing (basically UI is heaviest part in jmeter).
you can run your jmeter script in non-ui mode like this,
Jmeter -n -t -H proxy -P port
generally on a single dual core machine with 2 GB ram (Load Generator in your case) 100 user test can be carried out successfully.
some more things you can look at to find out the actual bottleneck
1.check application server logs (server on which your application is hosted)
if there are any failures in that then see performance counters on server (CPU, Memory, network etc) to see anything is overloaded.
(if server is windows then check using perfmon if linux then try sar)
if something is overloaded then reason is your app server cant take load of 100 users
probably try tuning it more.
2.check load generator system performance counters (JVM heap usage,CPU,Memory etc)
if JVM heap size is small enough try increasing it but if other counters are overloaded then try distributed load testing.
3.remove unwanted/heavy listeners, assertion from script.
maybe this will help :)
I want to run a Load test on my Production server to verify that the server can handle 1 million requests per 10 seconds using JMeter. How to configure JMeter Thread group for 1 million request in 10 seconds? How many client I need to do this test?
Please share your valuable experience if you have experience doing this type of load test.
First, you should ensure you really need 1 million requests in 10 seconds (what kind of site are you testing ?).
Then if you want to use JMeter, ensure:
You use last version
You tune memory correctly
You will certainly have to use distributed testing if not using Cloud
Follow best practices
http://www.dzone.com/links/r/see_how_to_make_jmeter_run_thousands_of_threads_w.html (Disclaimer : I'm the writer of this)
http://jmeter.apache.org/usermanual/best-practices.html
You might try Constant Throughput Timer to make a kind of Barrier
Alternatively you can try "Delay thread until creation"
And finally try a Cloud solution to get to this load, see this french blog on all kind of issues you will face in all fields (not only load software):
http://blog.milamberspace.net/index.php/2012/07/14/rapport-de-tres-gros-test-de-charge-avec-la-solution-blazemeter-1161.html)
But I never tried up to this load, so I cannot tell if it will work and it is kind of unexplored field
you can accomplish this using Gatling and scaling out to several machines that would run the test in parallel. in the end Gatling can aggregate the results into 1 report.
Gatling documentation dosent provide much info for this :
http://gatling.io/docs/1.5.6/user_documentation/cookbooks/scaling_out.html
but you can check my blog post to see how this can be done (ive wrote a script for this purpose)
Blog - http://www.nimrodstech.com/gatling-cluster-load-testing/
Gist - https://gist.github.com/Nimrod007/5cfed34eeffedfd7ec76
Looks like you'd better look onto another, more suitable tool for such kind of scenario,
e.g. Tsung
Tsung homepage
Tsung # github
Scaling to 30K: Tsung
or at least Gatling instead:
Gatling homepage
Gatling: scaling out
Perhaps if you would like to use jmeter in any case you can look onto BlazeMeterLoad Testing Cloud solution.