Ok then the sample I posted before *should* work +Request(ID1) ++Extract Id's into say fetchIds. +while (someVariableIsNotSet) ++For Each Id in fetchIds +++Request ++++Post Proc Extract the ids say nextIds ++++Listener combine nextIds into some suitable datastructure say allNextIds ++Beanshell sampler If there are some id's still be fetched, set the structure(allNextIds) into the form the FOR loop expects(fetchIds) otherwise set the variable causing the while to terminate
Do you know beanshell/java? otherwise the Listener step will be difficult to write. Note that the order in which your data will be fetched is ID1,ID2,ID3,ID1-1,ID1-2,ID2-1,ID2-2,ID2-3,... im assuming that the order is irrelevant to your test. regards deepak On Thu, Jan 14, 2010 at 12:57 PM, 7BOOK <[email protected]> wrote: > > sorry wrong example, you know, should be (unique nodeid) : > ID1 > ID1-1 > ID1-2 > ID2 > ID2-1 > ID2-1-1 > ID2-1-2 > ID2-2 > ID2-3 > ID2-3-1 > ID2-3-1-1 > ID2-3-1-2 > ... > ... > ID3 > ... > > > > 7BOOK wrote: > > > > Hi Deepak, > > > > sorry, i did not make it clear. > > yes, I just want to run this plan with single thread to track all ids (id > > unique, and never be requested again in any level), > > for example,just like browing window's explorer, try to open all the > > folders?: > > ID1 > > ID1-1 > > ID1-2 > > ID2 > > ID2-1 > > ID2-1 > > ID2-2 > > ID2-2 > > ID2-3 > > ID3-1 > > ID3-1-1 > > ID3-1-2 > > ... > > ... > > ID3 > > ... > > > > and the tree above is not infinite > > > > thanks Deepak > > > > > > > > Deepak Shetty wrote: > >> > >>>ID is unique, > >> Sorry i should have been clear. Say your first request extracts id > >> 1,2,3,4,...,9,10 > >> Say you request Page with id 1, is there a chance that you will get any > >> of > >> the previous ids (at any level) > >>>and my volumes are not large, > >> Ok. That should allow you to use in memory. > >>>so if I don't specify a loop count or something > >> But when do you terminate? when there are no more ids ? or is it meant > to > >> be > >> infinite? > >> > >> Also what about threads? Do you want to run this in a single thread? or > >> multiple threads? should each thread request the same set of ids or > >> different? > >> > >> > >> regards > >> deepak > >> > >> > >> > >> On Wed, Jan 13, 2010 at 9:26 PM, 7BOOK <[email protected]> wrote: > >> > >>> > >>> Hi Deepak, > >>> as you asked, ID is unique, and my volumes are not large, > >>> I've read spidering a site, so if I don't specify a loop count or > >>> something > >>> else which is fixed, > >>> waht's the best way for it? > >>> > >>> thanks! > >>> > >>> > >>> > >>> Deepak Shetty wrote: > >>> > > >>> > whats your terminating condition? are the id's always unique ? do you > >>> care > >>> > if you fetch the same id multiple times? what volumes are you talking > >>> > about > >>> > here? > >>> > > >>> > Your test looks like a sort of spider which isn't easy to do in > JMeter > >>> > > >>> > Thread group > >>> > +Request > >>> > ++Extract Id's into say fetchIds > >>> > +while (someVariableIsNotSet) > >>> > ++For Each Id in fetchIds > >>> > +++Request > >>> > ++++Post Proc Extract the ids say nextIds > >>> > ++++Listener combine nextIds into some suitable datastructure say > >>> > allNextIds > >>> > ++If there are some id's still be fetched, set the > >>> structure(allNextIds) > >>> > into the form the FOR loop expects(fetchIds) otherwise set the > >>> variable > >>> > causing the while to terminate > >>> > > >>> > If your volumes are large, you cant do this in memory. > >>> > See > >>> > > >>> > http://theworkaholic.blogspot.com/2009/10/spidering-site-with-jmeter.html#solution > >>> > for a sample on how to use files to do the above > >>> > > >>> > regards > >>> > deepak > >>> > > >>> > > >>> > > >>> > > >>> > > >>> > On Wed, Jan 13, 2010 at 8:57 PM, 7BOOK <[email protected]> wrote: > >>> > > >>> >> > >>> >> Hi all > >>> >> > >>> >> I am trying to make a jmeter plan to realize following regression > >>> loop: > >>> >> > >>> >> TestPlan > >>> >> Thread Group > >>> >> Http InitPage > >>> >> XPath ID > >>> >> For Each ID > >>> >> Http Request: Path:.../files/${ID} > >>> >> XPath ID > >>> >> For Each ID > >>> >> Http Request: Path:.../files/${ID} > >>> >> XPath ID > >>> >> For... > >>> >> ... > >>> >> > >>> >> ... > >>> >> > >>> >> The plan will stop until all related IDs have been found and > >>> requested. > >>> >> I can't use counter or variable to define a fixed looping depth > >>> level. > >>> >> I've tested Module Controller, from one module call back to it's > >>> caller, > >>> >> it > >>> >> caused stackoverflow,nothing happened, > >>> >> > >>> >> Thanks! > >>> >> -- > >>> >> View this message in context: > >>> >> > >>> > http://old.nabble.com/how-to-use-jmeter-plan-to-realize-jmeter-regression-loop-tp27156441p27156441.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] > >>> >> > >>> >> > >>> > > >>> > > >>> > >>> -- > >>> View this message in context: > >>> > http://old.nabble.com/how-to-use-jmeter-plan-to-realize-jmeter-regression-loop-tp27156441p27156594.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] > >>> > >>> > >> > >> > > > > > > -- > View this message in context: > http://old.nabble.com/how-to-use-jmeter-plan-to-realize-jmeter-regression-loop-tp27156441p27167975.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] > >
