There's something else at play then. As I mentioned in the earlier message,
your code example does not affect PK generation at all.
Andrus
On Jun 22, 2011, at 12:27 AM, Bruno René Santos wrote:
> That did not solve the problem... Maybe the use of shared cache is a
> problem? I turn off the cach
That did not solve the problem... Maybe the use of shared cache is a
problem? I turn off the cache alltogether and the problem did not happen
anymore.
Regards
Bruno
On Sun, Jun 19, 2011 at 9:54 AM, Andrus Adamchik wrote:
>
> On Jun 16, 2011, at 7:49 PM, Chris Poulsen wrote:
>
> >
> >
On Jun 16, 2011, at 7:49 PM, Chris Poulsen wrote:
>
>ORACLE
>PK_CONTACT_PERSON_KEY_SEQ
>1
>
Yep, that's the trick to map a sequence with increment of 1. On the Modeler
side it is configured here:
http://people.apache.org/~aadamchik/sequence.p
Cayenne PK generator caches blocks of 20 keys for each entity by default to
reduce the number of DB trips for such trivial
things as PKs.
> We think this is a bug because on the PK generation phase on cayenne no data,
> always, should be ever retrieved from the cache, which would lead to repea
Hi Bruno,
We've had issues at work related to caching of sequence values on oracle if
I recall correctly. We're using cayenne 3.
As we're not doing bulk inserts through cayenne, we decided to simply
extract one key at the time (not cache). The following xml snippet shows how
we configure the db-k
Hello all,
We had a problem on one of our applications on Oracle because, we think,
of the cayenne cache and we wanted to have some sort of confirmation if
we are seeing this correctly...
On Oracle, in order to get the next PK we need to execute the query:
Select sequence_name.next_val from
