Grouper Developers Forum

[Neil Matatall <>]

I would prefer the fatal error report.  If I recall correctly, we can 
run the auto-init target whenever we want.  Perhaps the fatal error 
report should have some message about running auto-init?

Chris Hyzer wrote:
> Sure, anyone let me know if they would prefer that...  fyi the auto-init 
> inserts the root stem if not there, several built in fields if not there, 2 
> built in types if not there...  it doesn't do much.
>
> Chris
>
> > -----Original Message-----
> > From: Tom Barton 
> > [mailto:]
> > Sent: Tuesday, September 02, 2008 3:36 PM
> > To: Grouper Dev
> > Subject: Re: [grouper-dev] improving db.init
> >
> > Thanks Chris. I wonder if some sites would prefer a fatal error report
> > rather than an auto-init upon detecting a corrupted database at
> > startup.
> > They might need the corrupted DB to aid in restoring normal operation.
> >
> > Tom
> >
> > Chris Hyzer wrote:
> > > This is done:
> > >
> > >
> > >
> > > https://bugs.internet2.edu/jira/browse/GRP-159
> > >
> > >
> > >
> > > We briefly discussed this one, but trying to make it easier to get
> > > going with grouper...
> > >
> > > Three db change tasks are confusing: schemaexport (ddl [and
> > optionally
> > > init data]), db.init (ddl then insert default data), db.reset (remove
> > > all data, then init).
> > >
> > > Now, when grouper starts, it checks to see if the root stem is
> > > missing, or if a builtin field is missing, or a builtin type type is
> > > missing, and if so, will init the DB. The only time this isnt
> > > sufficient is if some fields or types are there but not others (if
> > > someone manually deleted some), in which case the user can set the
> > > setting in the
> > > grouper.properties: ddlutils.schemaexport.installGrouperData to true,
> > > and do a schemaexport. I removed the ant tasks for db.init, db-init,
> > > and db.init.test, since they are no longer needed...
> > >
> > > If there is a place where grouper is starting without initting, just
> > > add this to the code before initted data is needed:
> > >
> > >     GrouperStartup.startup();
> > >
> > > Let me know if anyone has comments, I can always undo or change...
> > >
> > > [ Show > <https://bugs.internet2.edu/jira/browse/GRP-159> ]
> > >
> > > Chris Hyzer
> > <https://bugs.internet2.edu/jira/secure/ViewProfile.jspa?name=mchyzer>
> > > -
> > > [24/Aug/08 09:24 PM - edited ] fixed Now, when grouper starts, it
> > > checks to see if the root stem is missing, or if a builtin field is
> > > missing, or a builtin type type is missing, and if so, will init the
> > > DB. The only time this isnt sufficient is if some fields or types are
> > > there but not others (if someone manually deleted some), in which
> > case
> > > the user can set the setting in the grouper.properties:
> > > ddlutils.schemaexport.installGrouperData to true, and do a
> > schemaexport.
> > > I removed the ant tasks for db.init, db-init, and db.init.test, since
> > > they are no longer needed... If there is a place where grouper is
> > > starting without initting, just add this to the code before initted
> > data
> > > is needed:     GrouperStartup.startup(); Let me know if anyone has
> > > comments, I can always undo or change...
> > >
> > >
> > > Chris