I ran ab and Jmeter. For ab I used 200 concurrency with a total of 20000 requests and for Jmeter 200 sized thread group for 100 loops. The results of the #requests/sec varied a lot. From ab I get ~1000 and Jmeter gives me 500 at max. Also while setting verbosity of the responses for the ab test I receive very few (~100) responses back. Is it the reason why abTest is performing better?
Asked
Active
Viewed 526 times
0
-
This [thread](http://stackoverflow.com/questions/10260526/which-gets-the-measurements-right-jmeter-or-apache-ab) might help – Adnan Oct 02 '16 at 03:11
1 Answers
2
Are you following JMeter best practices when running your test:
If not, then it explains your issues.
Another possibility is that you're not setting a startup delay of 10 seconds for JMeter to create the threads, which impacts then your test.
UBIK LOAD PACK
- 33,980
- 5
- 71
- 116
-
sorry the values shown by the two tests are still coming out to be different. I'm sticking to ab right now. – fight2glory Oct 07 '16 at 05:28
-
What excatly did you compare in the 2 runs ? ab vs jmeter and how did you proceed ? – UBIK LOAD PACK Oct 07 '16 at 05:42
-
Im comparing the Request per second parameter. The configs I have mentioned in the question. – fight2glory Oct 07 '16 at 05:45
-
can you show jmeter test plan including thread group config detail and show ab command line ? thx – UBIK LOAD PACK Oct 07 '16 at 06:07
-
I am currently tied up with some other issue, but will update it soon. Thanks. – fight2glory Oct 07 '16 at 06:10