Re: RE dblink:

Thanks for pointing out that there is a paper on the oriole corp web site re: dblinks. I'll go check it out.

One quick question. Is a dblink created on the main repository side when OEM uses say (capacity planner or what not) Is that what happens when you save it as "preferred credentials"?

Just wondeirng how OEM is using the dblink and where.

Thanks again,

        Heidi

> dblinks are the only way with SQL*Plus to open several connections at
> once. Useful for comparing schemas between two instances, for example
> (select object_name from user_objects_at_here minus select object_name from
> user_objects_at_there), or data. Your idea of a global collector is quite
> valid - see OEM. I personally dislike GLOBAL_NAME=TRUE because I see no
> reason why the database link name should be the same as the instance
> name - from an applicative point of view, you may need to know that the
> data is managed by another application, but the instance or database
> name belongs to the nuts and bolts category. If you can avoid
> GLOBAL_NAME=TRUE (not always possible - if you are using replication for
> instance) ... If you are interested in the topic of naming dblinks,
> there is a paper in the library section of our site.
> A loopback dblink forces you to pass by (some) SQL*Net layers, so may
> be a way to test things. It is also a protection against environmental
> hazards, eg improperly set or unset ORACLE_SID, or set TWO_TASK.
>
> --
> Regards,
>
> Stephane Faroult
> email: sfaroult_at_oriolecorp.com
> Oriole Corporation
> ------------------------------------------------------------------
> http://www.oriolecorp.com, designed by Oracle DBAs for Oracle DBAs
> ------------------------------------------------------------------
Received on Fri Jun 02 2000 - 13:52:39 CDT