On 8 September 2010 11:55, Jens Müller <[email protected]> wrote: > > Hello, > it seems the CSV data set config does not correctly work > with distributed testing. > The semantic of the sharing > mode "all threads" is not as expected, it is only shared by the > local threads on one machine, as the files are read locally > separately.
Client server mode only transfers the test plan to the servers, and the servers return the test data. > It should be transferred together with the test plan > to all computers running it. If the same file were transferred to all the servers, then duplications would occur. If the client had to process reading the CSV file for all the servers, then it could place a large load on the client and network. > Any idea how I can achieve > this (no duplicate usage of the same elements of the file > during one test run)? Ensure that the CSV files are all unique. > Thanks > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > > --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
