You are viewing a plain text version of this content. The canonical link for it is here.
Posted to ojb-user@db.apache.org by Bonnie MacKellar <BM...@mobius.com> on 2003/05/07 23:01:56 UTC

correct sequence manager with MSSqlServer and Weblogic

We are trying to run OJB with MSSqlServer and Weblogic. We are NOT
using EJBs, however, just JNDI DataSources.

After seeing the post earlier today about not using the High/Low sequence
manager
implementation with datasources, I went to the documentation, and sure
enough, it
does say that. I missed it earlier. Unfortunately, I have been using it.
It looks like the database-based implementations are out because we are
using
MSSqlServer. So that leaves the in-memory sequence manager. The
documentation
just says not to use in a clustered environment. Does anyone know if it is
safe with managed datasources? Or does it have problems like the High/Low
manager? By the way, our database will not be updated by other applications.

thanks,
Bonnie MacKellar
software engineer
Mobius Management Systems, Inc.
bmackell@mobius.com


Re: correct sequence manager with MSSqlServer and Weblogic

Posted by Armin Waibel <ar...@code-au-lait.de>.
Hi Bonnie,

> We are trying to run OJB with MSSqlServer and Weblogic. We are NOT
> using EJBs, however, just JNDI DataSources.
>
> After seeing the post earlier today about not using the High/Low
sequence
> manager
> implementation with datasources, I went to the documentation, and sure
> enough, it
> does say that. I missed it earlier. Unfortunately, I have been using
it.
> It looks like the database-based implementations are out because we
are
> using
> MSSqlServer. So that leaves the in-memory sequence manager. The
> documentation
> just says not to use in a clustered environment. Does anyone know if
it is
> safe with managed datasources? Or does it have problems like the
High/Low
> manager? By the way, our database will not be updated by other
applications.

Should be safe in your case (and definitely the best performing
solution).
The in-memory implementation is 'nearly independent' from the
database. Only the first time an UID is needed, in the associated
db-table the max id of the PK-column was calculated using a
db-select. All following UID calls only increment the calculated
value.

Using MSSQL you could use another alternative implementation,
please see

http://db.apache.org/ojb/sequencemanager.html#Oracle-style sequencing

But I think in your specific case the in-memory implementation
is the best solution.

regards,
Armin

>
> thanks,
> Bonnie MacKellar
> software engineer
> Mobius Management Systems, Inc.
> bmackell@mobius.com
>
>