Skip to Content.
Sympa Menu

grouper-users - RE: [grouper-users] Grouper 1.4.0: Unable to delete broken stem

Subject: Grouper Users - Open Discussion List

List archive

RE: [grouper-users] Grouper 1.4.0: Unable to delete broken stem


Chronological Thread 
  • From: Chris Hyzer <>
  • To: "" <>
  • Subject: RE: [grouper-users] Grouper 1.4.0: Unable to delete broken stem
  • Date: Mon, 19 Jan 2009 12:10:26 -0500
  • Accept-language: en-US
  • Acceptlanguage: en-US

Couple more things:

1. When I moved the group name attributes to grouper_groups table, I made
unique indexes on group name and group display name. Maybe grouper_stems
needs unique constraints on stem_name and stem display name.

2.

> > > but I don't really understand the consequences of using
> > > READ_WRITE_OR_USE_EXISTING as opposed to READ_WRITE_NEW.

In a web service these are equivalent. If you were in Java, and you did
READ_WRITE_OR_USE_EXISTING, then it would detect if it were already in a
transaction context, if so use it, if not, make a new one which is
READ_WRITE. If you did READ_WRITE_NEW, then even if there were an existing
transaction, it would create a new autonomous one which didn’t interact with
the outer one

3. In a web service, if there is one operation (e.g. not batched), or if it
is batched with size one, then you don’t need to pass any transaction type.
Even if the operation touches many underlying tables (e.g. adding a stem will
also create some membership records for who is allowed to edit it), all that
will happen in one tx no matter what. However if it is batched, if you pass
READ_WRITE_NEW, then all in the batch will be one tx. If not passed, or if
NONE, then each individual operation will be its own transaction.

Is the Javadoc for 1.4 posted??? In any case, it should be in the binary
release, or you can build it with ant. Look at GrouperTransactionType
javadoc:

Here is source which is next best thing :)

http://viewvc.internet2.edu/viewvc.py/grouper/src/grouper/edu/internet2/middleware/grouper/hibernate/GrouperTransactionType.java?root=I2MI&view=markup

Regards,
Chris




> -----Original Message-----
> From: Chris Hyzer
> Sent: Monday, January 19, 2009 9:49 AM
> To:
>
> Subject: FW: [grouper-users] Grouper 1.4.0: Unable to delete broken
> stem
>
> fyi
>
> > -----Original Message-----
> > From: Chris Hyzer
> > Sent: Monday, January 19, 2009 9:49 AM
> > To:
> > ''
> > Subject: RE: [grouper-users] Grouper 1.4.0: Unable to delete broken
> > stem
> >
> > Sorry about that, this sounds like a bug. I will try to reproduce
> this
> > and fix it for 1.4.1.
> >
> > In the meantime, if you can delete all those from the DB (except one,
> > or all and recreate).
> >
> > Once you get your DB to a state where there are no dupes, the index
> on
> > stem name could be unique I believe... until 1.4.1 you might want to
> > check to see if a stem exists before creating (if that is the
> problem,
> > create twice without checking?) If you get more info about the web
> > service you call and params used to create multiples, let me know (I
> > assume its "save stem")
> >
> > https://bugs.internet2.edu/jira/browse/GRP-204
> >
> > Anyone know of a reason why stem name index shouldn’t be unique? I
> > don’t think it was unique in 1.3 (with hibernate schema export)
> > either...
> >
> > Thanks!
> > Chris
> >
> > > -----Original Message-----
> > > From: Dr. Loris Bennett
> > > [mailto:]
> > > Sent: Monday, January 19, 2009 4:39 AM
> > > To: Chris Hyzer
> > > Subject: RE: [grouper-users] Grouper 1.4.0: Unable to delete broken
> > > stem
> > >
> > > Hi Chris,
> > >
> > > Here is the log file. The offending stem "test:new_junk" has been
> > > created multiple times in the table "grouper_stems".
> > >
> > > I have been using GcTransactionType.NONE, which strikes me as
> unwise,
> > > but I don't really understand the consequences of using
> > > READ_WRITE_OR_USE_EXISTING as opposed to READ_WRITE_NEW.
> > >
> > > Cheers
> > >
> > > Loris
> > >
> > >
> > > On Fri, 2009-01-16 at 10:17 -0500, Chris Hyzer wrote:
> > > > Its hard to tell without more info... are there any error
> messages
> > in
> > > the logs?
> > > >
> > > > Quick and dirty might just a DB query: delete from grouper_stems
> > > > where name = "a:b:c";
> > > >
> > > > (then if you get foreign key problems with group records or
> > > something,
> > > > delete that stuff too... but it would be nice to figure out what
> > the
> > > > real problem is...)
> > > >
> > > > Regards,
> > > > Chris
> > > >
> > > >
> > > > > -----Original Message-----
> > > > > From: Dr. Loris Bennett
> > > > > [mailto:]
> > > > > Sent: Friday, January 16, 2009 8:22 AM
> > > > > To: Grouper Users Mailing List
> > > > > Subject: [grouper-users] Grouper 1.4.0: Unable to delete broken
> > > stem
> > > > >
> > > > > Hi,
> > > > >
> > > > > Through fiddling around with grouper client and possibly due to
> a
> > > > > version mismatch between client and web service I have managed
> to
> > > > > create a stem which I cannot delete via the UI or via grouper
> > > client.
> > > > > Also, when I click on it in the UI I reach a sort of dead end
> in
> > > > > which I can no longer browse by click on "explore" (see
> > > attachment).
> > > > >
> > > > > Any ideas how I can get rid of the stem?
> > > > >
> > > > > Loris
> > > > > --
> > > > > Dr. Loris Bennett (Mr.)
> > > > > Freie Universität Berlin
> > > > > ZEDAT - Zentraleinrichtung für Datenverarbeitung / Computer
> > Center
> > > > > Compute & Media Service Fabeckstr. 32, Room 221
> > > > > D-14195 Berlin
> > > > > Tel ++49 30 838 51024
> > > > > Fax ++49 30 838 56721
> > > > > Email
> > > > >
> > > > > Web www.zedat.fu-berlin.de
> > > --
> > > Dr. Loris Bennett (Mr.)
> > > Freie Universität Berlin
> > > ZEDAT - Zentraleinrichtung für Datenverarbeitung / Computer Center
> > > Compute & Media Service Fabeckstr. 32, Room 221
> > > D-14195 Berlin
> > > Tel ++49 30 838 51024
> > > Fax ++49 30 838 56721
> > > Email
> > >
> > > Web www.zedat.fu-berlin.de



Archive powered by MHonArc 2.6.16.

Top of Page