[Mondrian] DynamicDatasourcesXmlaServlet

Patrick Leckey patl at seewind.com
Mon Jul 18 10:34:50 EDT 2011


Luc & Julian,

Thanks, I'll update our testing environment to 3.3-SNAPSHOT.  We can't use 3.2.1.13885 as it has a problem with some NON EMPTY queries (IIRC, MONDRIAN-942 but that's just off the top of my head) which seems to have been fixed in the 3.2-SNAPSHOT.  We're rolling out 1.0 in early August, and that was a show-stopper for us, so we're playing with all the dev builds to see which is the best to go to production with.

So far 4.0.0.13695 seems to be the lead horse, and the same change fixes the DynamicDatasourcesXmlaServlet in that build too.

Other than the fact that's not an officially stable release, and assuming we're willing to take our chances, any reason we shouldn't use that build over the current 3.3 builds?

Thanks,
Pat

On 2011-07-16, at 10:26 AM, Luc Boudreau wrote:

> 
> Patrick,
> 
> Thanks for reporting this. I have fixed the issue and it will be resolved as of change 14462. It will be included in our next release, Mondrian 3.3.0.
> 
> As a side note, the 3.2-SNAPSHOT builds are from a dead branch that we do not maintain nor test anymore. There might be some weird stuff in there. I suggest switching to either 3.3-SNAPSHOT or stick to 3.2.1.13885 to prevent any problems.
> 
> Cheers!
> 
> Luc
> 
> On Fri, Jul 15, 2011 at 7:15 PM, Julian Hyde <jhyde at pentaho.com> wrote:
> An oversight. (We're human too!)
> 
> Please contribute your fix. And if you contribute a test case, we can keep
> it working in the future...
> 
> Julian
> 
> > -----Original Message-----
> > From: mondrian-bounces at pentaho.org
> > [mailto:mondrian-bounces at pentaho.org] On Behalf Of Patrick Leckey
> > Sent: Friday, July 15, 2011 12:10 PM
> > To: Mondrian developer mailing list
> > Subject: [Mondrian] DynamicDatasourcesXmlaServlet
> >
> > Hi,
> >
> > We're working with the latest 3.2-SNAPSHOT in a testing
> > environment and noticed that the
> > DynamicDatasourcesXmlaServlet no longer works, while the one
> > in 3.2.1.13885 did.
> >
> > Delving into the source, a lot has changed in this area.
> > Background scheduling the updates so it doesn't use up time
> > on each request - tres cool.  Hugely useful for us.
> >
> > However, I noticed in both DynamicContentFinder and
> > FileRepository, where the scheduled executions are setup,
> > both thread pools are being initialized with a pool size of 0
> > - so the background threads never actually get executed, and
> > thus the datasource never gets checked for changes.
> >
> > In our testing environment I set both thread pools to
> > initialize the pool size to 1 instead, and now everything
> > seems to work beautifully.
> >
> > I'm just curious if this functionality was disabled for a
> > reason I haven't hit yet, or if this was merely an oversight?
> >
> > Testing environment is:
> > - CentOS 5.6
> > - Tomcat 5.5.23
> > - OpenJDK 1.6_21
> >
> > Pat
> > _______________________________________________
> > Mondrian mailing list
> > Mondrian at pentaho.org
> > http://lists.pentaho.org/mailman/listinfo/mondrian
> >
> 
> _______________________________________________
> Mondrian mailing list
> Mondrian at pentaho.org
> http://lists.pentaho.org/mailman/listinfo/mondrian
> 
> _______________________________________________
> Mondrian mailing list
> Mondrian at pentaho.org
> http://lists.pentaho.org/mailman/listinfo/mondrian

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.pentaho.org/pipermail/mondrian/attachments/20110718/dcdf6cd3/attachment.html 


More information about the Mondrian mailing list