Thanks for the tips as always... I will try this, but do you think this is a "hack" 
since
this (sequence generation) is not what keys are intended for?  See my other post please
(coming soon).

"we welcome additions to the documentation."

This is truly open-source. :)  I will try to throw my contribution in at some point; 
just
too busy trying to meet deadlines (aren't we all?)

Nate

John McNally wrote:

> you can set
> database.idbroker.cleverquantity=false
> and then set the Quantity=1 in ID_TABLE for tables which the skipped
> pk's are an issue.
>
> this has the same effect as the prefetch property, but is more
> granular.  The prefetch property has been around for a while, but maybe
> it is not in the released version.
>
>
>
> john mcnally
>
> Nathaniel Reed wrote:
> >
> > Any other ideas??
> >
> > I found "" (which I changed to false, & I don't
> > think there is a "DatabaseService") in TR.props, so I added
> >
> > database.idbroker.prefetch=false
> >
> > but the problem I describe persists.  I think this is a non-existant property as 
>far
> > as Turbine is concerned -- are you sure it is used and if so, which class uses it?
> >
> > I was scanning the code (IDBroker.java) and cannot determine how the IDBroker is
> > configured, since the only apparent access is through method calls from TableMap,
> > which is accessed through DatabaseMap, which is accessed by TurbineDB, a service
> > which uses the "database.<propname>" properties for obtaining JDBC connections in
> > TR.props.  Other than the connection info, I cannot determine how the TR.props are
> > used by TurbineDB.
> >
> > All these wonderfully sophisticated design patterns obscure the most basic
> > information about how to configure basic Turbine services/utilities.  I would think
> > should be a top priority in the documentation...
> >
> > Nate
> >
> > John McNally wrote:
> >
> > > idbroker grabs a series of ids and keeps the count in memory until the
> > > range is used up.  If all the ids are not handed out before the app is
> > > stopped there will be gaps.  There is no attempt to keep these
> > > sequential as this should be unimportant in a pk.  However if you do not
> > > want to cache you can use
> > >
> > > <services.DatabaseService>.<database.>idbroker.prefetch=false
> > >
> > > and id's will be grabbed from the db one at a time on an as needed
> > > basis.  the stuff in <> are guesses and/or optional, look for it (or one
> > > like it) in TR.props.
> > >
> > > john mcnally
> > >
> > > Nathaniel Reed wrote:
> > > >
> > > > I'm using Turbine 2.1 on the Ingres RDBMS.
> > > >
> > > > Is there a way to force the ID broker to generate sequential keys for my
> > > > table?  For some reason, it "jumps" to a higher number whenever I
> > > > restart the application and then create another object.  If the sequence
> > > > starts at 100, for instance, I create 100, 101, and 102, then if I
> > > > restart the next object I create has id 120.  What's going on?
> > > >
> > > > Nate
> > > >
> > > > --
> > > > Nate Reed
> > > > Physical Oceanography Distributed Active Archive Center
> > > > Jet Propulsion Laboratory (Raytheon)
> > > > [EMAIL PROTECTED]
> > > > (626) 744-5528
> > > > (626) 744-5506
> > > >
> > > > --
> > > > To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
> > > > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
> > >
> > > --
> > > To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
> > > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
> >
> > --
> > Nate Reed
> > Physical Oceanography Distributed Active Archive Center
> > Jet Propulsion Laboratory (Raytheon)
> > [EMAIL PROTECTED]
> > (626) 744-5528
> > (626) 744-5506
> >
> > --
> > To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
> > For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>
>
> --
> To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
> For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

--
Nate Reed
Physical Oceanography Distributed Active Archive Center
Jet Propulsion Laboratory (Raytheon)
[EMAIL PROTECTED]
(626) 744-5528
(626) 744-5506



--
To unsubscribe, e-mail:   <mailto:[EMAIL PROTECTED]>
For additional commands, e-mail: <mailto:[EMAIL PROTECTED]>

Reply via email to