Yep, I did extraction from JDBC sampler and put it into CSV with "Save response to a file" Listener + a post processor to remove first line:
//Removing first line (column names) of results String result; result = prev.getResponseDataAsString(); removeString = "CONCAT(name, '/', id)\n"; rlen = removeString.length(); result = result.substring(rlen,result.length() - rlen); prev.setResponseData(result); Best, Shmuel. On Thu, Jun 21, 2012 at 7:03 PM, sebb <[email protected]> wrote: > On 21 June 2012 16:27, Roderick Parks <[email protected]> > wrote: > > I've recently developed a prototype JDBC Data Set component that is > > analogous to the CSV Data Set except that the values come from the > > columns in the select or prepared select statement. It uses JDBC > > Connection Configuration just like the JDBC sampler, pre-processor and > > post-processor. > > It would be likely more efficient to extract the variables into a CSV > file just before running the test. > Of course that may not be quite as convenient, and would not work if > the values are dynamic. > > > Early on I thought that keep alive and connection age could be > > significant parameters affecting the performance and efficiency of my > > component but when I looked in the source code for the JDBC Connection > > (src/protocol/jdbc/org/apache/jmeter/protocol/jdbc/config/DataSourceElem > > ent.java) I found not only that "Keep-Alive", "Max Connection Age" and > > "Validation Query" have set/get methods but are otherwise not used > > (presumably an aspirational addition to the component and its beaninfo) > > but also that it uses the now defunct Apache Excalibur as its actual > > JDBC interface: it's only a wrapper. Rather than dig into another > > source project, and a dead and static one at that, I decided to trust > > JMeter's JDBC Connection Configuration's encapsulation of Excalibur and > > hope that the eventually inevitable migration to something else remains > > equally encapsulated and won't break my component. > > > > So that's the long answer. The short answer is that you have no control > > over the status of a given connection and will have to tolerate the > > default behaviour. > > > > -----Original Message----- > > From: Dzmitry_Kashlach [mailto:[email protected]] > > Sent: 21 June 2012 15:08 > > To: [email protected] > > Subject: Re: How to connect to Database from JMeter at the start of the > > test and use the connection in the post processor to insert some data? > > > > When I had task that was similar to yours, I opened connection to > > database > > via JDBC Connection Configuration once per Thread Group. After that I > > used > > this connection through the whole test. While creating > > JMeter test-plan, I used > > http://community.blazemeter.com/knowledgebase/articles/65143-using-jdbc- > > sampler-in-jmeter-2-6 > > How to use JDBC Sampler in JMeter > > > > -- > > View this message in context: > > http://jmeter.512774.n5.nabble.com/How-to-connect-to-Database-from-JMete > > r-at-the-start-of-the-test-and-use-the-connection-in-the-post-p-tp567638 > > 8p5713661.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] > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: [email protected] > For additional commands, e-mail: [email protected] > >
