hello,
please I would like to know if it's possible to load data in
hierarchical table structure with identity column used as pk (for all
the tables) :
I mean

table root (Id_root(pk,identity column generated always), att_r1,...)
table leaf (Id_leaf(pk,identity column generated always), id_root(fk
ref root), att_l1 ...)

could I use import to populate both root and leaf table avoiding an
insert select insert loop ?

many tanks in advance
best regard David

If you reset the identity in all the tables and if you are doing a load
with replace then you might end up with the same identity values
everywhere but only if there are no rejects.

davvid@libero.it wrote:

> hello,
> please I would like to know if it's possible to load data in
> hierarchical table structure with identity column used as pk (for all
> the tables) :
> I mean
>
> table root (Id_root(pk,identity column generated always), att_r1,...)
> table leaf (Id_leaf(pk,identity column generated always), id_root(fk
> ref root), att_l1 ...)
>
> could I use import to populate both root and leaf table avoiding an
> insert select insert loop ?
>
> many tanks in advance
> best regard David

--
Anton Versteeg
IBM Certified DB2 Specialist
IBM Netherlands

thank you Anton but I didn't understand:
once the root table has been loaded how is possible the leaf table
knows every row's correct id_root to be loaded?

I'dont think is possible use the load utility.
I've read about using import to accomplish that. dont know how.

Anton Versteeg <anton_versteeg@nnll.ibm.com> wrote in message news:<c7dj9h$4n1c$1@sp15en20.hursley.ibm.com>...
> If you reset the identity in all the tables and if you are doing a load
> with replace then you might end up with the same identity values
> everywhere but only if there are no rejects.
>
> davvid@libero.it wrote:
>
> > hello,
> > please I would like to know if it's possible to load data in
> > hierarchical table structure with identity column used as pk (for all
> > the tables) :
> > I mean
> >
> > table root (Id_root(pk,identity column generated always), att_r1,...)
> > table leaf (Id_leaf(pk,identity column generated always), id_root(fk
> > ref root), att_l1 ...)
> >
> > could I use import to populate both root and leaf table avoiding an
> > insert select insert loop ?
> >
> > many tanks in advance
> > best regard David

Apparantly you have established some link between the tables.
I assumed it was the physical sequence of the load/import files.
If that is not the case then you should probably consider using
'generated by default' in stead.
In that case the identity values can be part of the load/import file.

davvid@libero.it wrote:
> thank you Anton but I didn't understand:
> once the root table has been loaded how is possible the leaf table
> knows every row's correct id_root to be loaded?
>
> I'dont think is possible use the load utility.
> I've read about using import to accomplish that. dont know how.
>
> Anton Versteeg <anton_versteeg@nnll.ibm.com> wrote in message news:<c7dj9h$4n1c$1@sp15en20.hursley.ibm.com>...
>
>>If you reset the identity in all the tables and if you are doing a load
>>with replace then you might end up with the same identity values
>>everywhere but only if there are no rejects.
>>
>>davvid@libero.it wrote:
>>
>>
>>>hello,
>>>please I would like to know if it's possible to load data in
>>>hierarchical table structure with identity column used as pk (for all
>>>the tables) :
>>>I mean
>>>
>>>table root (Id_root(pk,identity column generated always), att_r1,...)
>>>table leaf (Id_leaf(pk,identity column generated always), id_root(fk
>>>ref root), att_l1 ...)
>>>
>>>could I use import to populate both root and leaf table avoiding an
>>>insert select insert loop ?
>>>
>>>many tanks in advance
>>>best regard David

--
Anton Versteeg
IBM Certified DB2 Specialist
IBM Netherlands