We are executing SQL queries on Jmeter JDBC request and able to get proper response on executions. However, our project team would like to see row count of each execution in result reports along with elapsed time, latency etc.
How to get row count for a sql query executed in JDBC request in Jmeter? Also, how to get the row count value appended to columns in Summary report.csv file?
With regards to row count - you already have it. I.e. in the below setup I defined the result variable myVar
When my query is being executed JMeter stores the number of returned rows into myVar_# variable
Now if I want to display the variable within the .jtl results file I need to add the following line to user.properties file:
sample_variables=myVar_#
Next time I run my JMeter test in command-line non-gui mode like:
jmeter -n -t test.jmx -l result.csv
I see the number of rows returned by the SQL query in the myVar_# column:
References:
Sample Variables
Debugging JDBC Sampler Results in JMeter
Related
We are running performance tests using JMeter. In the Aggregate report we are able to view the minimum and maximum times. We now want to identify exactly which request from JMeter took the maximum time that was reported, so that we can perform a root cause analysis. So, how do we identify that request in JMeter?
If you run JMeter in command-line non-GUI mode like
jmeter -n -t test.jmx -l result.csv
it will generate CSV file containing all samplers statistics so if you sort by elapsed column in the timeStamp column using MS Excel or equivalent you will see the Unix timestamp so you can figure out the exact time when the "longest" request was fired:
Another option is adding a unique identifier like number of current thread, number of iteration, URL, whatever
it can be done using JMeter Functions or Pre-Defined Variables
Trying to compare the data/count between source and destination tables (SQL) as below through using CSV config in JMeter. is it possible to implement this way?.
CSV config in JMeter
just want to compare the count between STAGING and DW for all tables. any thoughts on this..?
Just define Variable Names so JDBC Request sampler could store the query execution output into a JMeter Variable like:
In first JDBC Request sampler define "Variable Name" as result1 (replace "SQL Query" with the relevant JMeter Variable name holding the value from the CSV file)
In second JDBC Request sampler define "Variable Name" as result2
That's it, now you will have the count of rows returned from the first JDBC Request sampler as result1_1 and from the second as result2_1
Now you can compare the values using Response Assertion configured like:
in case of mismatch the relevant JDBC Request sampler will be marked as failed and you will be informed about the differences:
More information: Debugging JDBC Sampler Results in JMeter
I have a JDBC test suite that I run with JMeter, and I want to run that over many user accounts.
I'm doing that with a CSV dataset that includes all the accountIds I want to test, then I inject the accountId into my SQL queries to run the queries over all of the accounts I want to test (with ${ACCOUNTID}).
It's working great, the queries are actually done over all the accounts I want to test.
However, I would like to know the response time per account. Currently I'm not getting the accountId back in the result, so I can't filter my results by account. Is there a way to get that variable value to the JTL?
You can either run JMeter as follows:
jmeter -Jsample_variables=ACCOUNTID -n -t test.jmx -l result.jtl
Or add the next line to user.properties file to make the change permanent
sample_variables=ACCOUNTID
In both cases you will get an extra column in .jtl results file having ACCOUNTID variable value for each and every request.
More information:
Sample Variables
JMeter Properties Reference
Apache JMeter Properties Customization Guide
You can add variable name to Sampler name so it will aggregate requests per account id. Set Sampler as:
My Sampler ${ACCOUNTID}
Is it possible to pass query through variable? What I need is:
1. Get multiple queries from csv file.
2. Execute them one by one.
3. Store results
Yes you can do it by using CSV Data Set config and Bean shell post processor.
Step 1: Need to read queries from CSV file.
Step 2: Use JDBC request to query the database. I'm assuming that you already created database JDBC pool with the help of JDBC Connection Configuration
Step 3: Use Bean Shell post processor to save the results into a file.
To read queries from the file one by one you can use __StringFromFile() function like:
To store the output you can go for different options,
Save Responses to a file listener if query result is a simple/singe value
If your query result is more complex, i.e. you need to iterate the ResultSet object - you will need to do some scripting, check out Debugging JDBC Sampler Results in JMeter
If your query generates several JMeter Variables you can get them stored into .jtl results file via Sample Variables property.
In Jmeter, when I set the file name in the Filename text box, configure the results and run the tests, I see that the file does not have Min, Max, Avg Reponse Times, which are displayed in the summary report.
How can be these values added to report in Jmeter?
Or is there any plug-in which I can use to do that?
I believe JMeterPluginsCMD at http://code.google.com/p/jmeter-plugins/wiki/JMeterPluginsCMD can save Aggregate reports that contain all the data you see in the UI (like min/max/avg). It only saves in CSV, afaik