cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Hunsberger, Peter" <Peter.Hunsber...@stjude.org>
Subject RE: Are flowscript functions reentrant?
Date Tue, 03 Feb 2004 16:28:03 GMT
Christopher Oliver <res1cf5x@verizon.net> writes:

> >>>Unfortunately, the reentrancy problem is NOT fixed (sigh)....
> >>> 
> >>>
> >>>      
> >>>
> >>What error do you get?
> >>    
> >>
> >
> >Same as before; last request from a SendPage* into the pipeline 
> >clobbers the previous requests if they are running on the samepline.
> >
> >To recreate build a long running pipeline (we're returning 
> about 1,000 
> >rows out of a possible 9,000,000 rows in our case) and call it twice 
> >with different request parameters that are supposed to 
> return different
> >results:  Eg; a generator that wraps request generator and 
> sleeps for 
> >30 seconds or so before returning.
> >  
> >
> The problem I fixed in 2.1.4-dev was with aggregating 
> multiple pipelines 
> that used <map:call function>.  I don't know what's causing 
> the problem 
> in your case. If I have time I'll try to recreate and debug it.

We do use aggregation in the pipeline; a single pipeline is called from
the flow script and aggregates the data from a bunch of generators, but
I take it that, that is not this issue you where referring to?
 
I've traced this down to a single generator that is not behaving in a
reentrant fashion.  The generator simply calls an EJB then walks a
resulting collection to spit out a (long) linear list of results. The
request data is different going into the generator but the EJB data is
the same coming back. A cursory inspection shows no obvious reason why;
the code sets a couple of instance variables and then goes off and calls
the EJB. 

Anyone know if it is possible that when a generator is pulled from a
pool that the one passed back to handle the generate method might not be
the same one as called for the setup method?  So far there seems like a
vague possibility that could introduce this behavior though, if so, it's
a little convoluted?




Mime
View raw message