I have a .jmx file that I’ve been using to perform tests via command line for multiple users against a test server.
The command that I’ve been using to perform the tests is: Jmeter –n –t BaselineThroughput.jmx –Jusers=# -l log#.jtl Where # is the number of simultaneous users I want for the test. The test is designed to run for 5 minutes and then exit. This works fine and results in a log file where I can distinguish the users results by tn=”Thread Group 1-#” where # ranges from 1 to the number of users requested. I am now running a distributed test with 2 engines running on separate servers and the same setup, except that I start the jmeter-server on each of the remote clients and then execute with: Jmeter –n –t BaselineThroughput.jmx –R ###.###.###.###,###.###.###.### –Jusers=# -l log#.jtl Where I'm using real IP addresses (removed here for anonymity) and # users ranges from 1-5000. This appears to work fine, results collected, etc, except every entry in the log file is stored as tn=”Thread Group 1-1”, I would have expected tn=”Thread Group x-y” where “x” is the remote client number and “y” is the user # on the remote client. I am also seeing many less samples than I would have expected during the 5 minutes the test runs, close to the number I would expect with only one user. It is almost as if only one user is being run, or, all the other sample data is being dropped on the floor. Is this working correctly? Have I missed something? Is there a way to confirm which data was collected from which client? Thanks, Kirin, SF - California -- View this message in context: http://www.nabble.com/Question-regarding-user-logging-in-distributed-tests...-tp21985780p21985780.html Sent from the JMeter - User mailing list archive at Nabble.com. --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
