Modeler Question....

[Tony Giaccone <to...@giaccone.org>]

I'm trying to re-engineer a schema. I got through the process for a
postgres database and it works like a charm. Create the datamap, configure
the connection, refresh the db schema, and up pops the tables from the
database, just as you would expect.

I follow up with a similar task only this time trying to connect to an
Oracle DB and the process seems to be working up to the point where the
database schema should reflect the tables in the database, and there
nothing happens. The "Refresh DB Schema" button stays dim and inactive
almost as if it was stuck trying to pull back the schema. Now this is a
large db with a lot of tables and a lot of data in the tables (not that it
should matter, but...).  I've used the test button on the edit data source
preferences and it responds with a successful connection mesg.

I thought maybe it was just taking a long time, but i've waited 5 minutes
and nothing happens. Is there anyway to increase the logging level on the
modeler so that I can see what if anything is happening as it tries to load
up the schemal from this oracle db?



Tony

Re: Modeler Question....

[Andrus Adamchik <aa...@gmail.com>]

Hi Tony,

Yeah I remember how slow Oracle tests always were due to the schema operations. And I've also observed your problem even on databases with much faster metadata retrieval. So I agree that DB import process should attempt to apply DB-side metadata filtering to the maximum extent possible. JDBC spec does not allow the degree of filtering that Cayenne DB import does, but we can still trim the scope of retrieved metadata significantly. We discussed it with Nikita some time back, and we'll need to revisit. 

Andrus


> On Apr 29, 2021, at 10:37 AM, Tony Giaccone <to...@giaccone.org> wrote:
> 
> So here's a response to my own question. This process in the Oracle
> database takes so long because there are 15 different schemas each with
> potentially dozens and in some cases over 100 tables.  The result is an
> import which just takes a long long time. It might be nice if it were
> possible to limit the scope of the data import to the schema defined in the
> left panel.  That would give an easy way to keep the scope of what's going
> on down to something more manageable.  If there were no schema's defined
> there, then fine pull everything.  Just a thought.
> 
> 
> Tony
> 
> On Wed, Apr 28, 2021 at 12:47 PM Tony Giaccone <to...@giaccone.org> wrote:
> 
>> 
>> 
>> I'm trying to re-engineer a schema. I got through the process for a
>> postgres database and it works like a charm. Create the datamap, configure
>> the connection, refresh the db schema, and up pops the tables from the
>> database, just as you would expect.
>> 
>> I follow up with a similar task only this time trying to connect to an
>> Oracle DB and the process seems to be working up to the point where the
>> database schema should reflect the tables in the database, and there
>> nothing happens. The "Refresh DB Schema" button stays dim and inactive
>> almost as if it was stuck trying to pull back the schema. Now this is a
>> large db with a lot of tables and a lot of data in the tables (not that it
>> should matter, but...).  I've used the test button on the edit data source
>> preferences and it responds with a successful connection mesg.
>> 
>> I thought maybe it was just taking a long time, but i've waited 5 minutes
>> and nothing happens. Is there anyway to increase the logging level on the
>> modeler so that I can see what if anything is happening as it tries to load
>> up the schemal from this oracle db?
>> 
>> 
>> 
>> Tony
>> 

Re: Modeler Question....

[Tony Giaccone <to...@giaccone.org>]

So here's a response to my own question. This process in the Oracle
database takes so long because there are 15 different schemas each with
potentially dozens and in some cases over 100 tables.  The result is an
import which just takes a long long time. It might be nice if it were
possible to limit the scope of the data import to the schema defined in the
left panel.  That would give an easy way to keep the scope of what's going
on down to something more manageable.  If there were no schema's defined
there, then fine pull everything.  Just a thought.


Tony

On Wed, Apr 28, 2021 at 12:47 PM Tony Giaccone <to...@giaccone.org> wrote:

>
>
> I'm trying to re-engineer a schema. I got through the process for a
> postgres database and it works like a charm. Create the datamap, configure
> the connection, refresh the db schema, and up pops the tables from the
> database, just as you would expect.
>
> I follow up with a similar task only this time trying to connect to an
> Oracle DB and the process seems to be working up to the point where the
> database schema should reflect the tables in the database, and there
> nothing happens. The "Refresh DB Schema" button stays dim and inactive
> almost as if it was stuck trying to pull back the schema. Now this is a
> large db with a lot of tables and a lot of data in the tables (not that it
> should matter, but...).  I've used the test button on the edit data source
> preferences and it responds with a successful connection mesg.
>
> I thought maybe it was just taking a long time, but i've waited 5 minutes
> and nothing happens. Is there anyway to increase the logging level on the
> modeler so that I can see what if anything is happening as it tries to load
> up the schemal from this oracle db?
>
>
>
> Tony
>