> *changing this to “false”
While this may solve the immediate problem, turning off shared object cache is
not terribly efficient. I would recommend against it.
>>> weirdly I see one of the
>>> relationship queries fire again, but not the other…either way both objects
>>> “snapshots” (EOF term)
Hello Andrus,
Thanks for trying to help out. I'm using MariaDB
(10.1.12-MariaDB-1~trusty), (org.mariadb.jdbc.Driver, JAR file
mariadb-java-client-1.2.0.jar).
I can check Cayenne version 4.0.M3 also.
/Fredrik
Here's the stored procedure I try to call.
CREATE DEFINER=`foodbase_admin`@`localhost`
*changing this to “false”
On Tue, Apr 19, 2016 at 4:07 PM, Lon Varscsak
wrote:
> It looks like I had cayenne.DataDomain.sharedCache=true in my model.
> Changing this to default solved my issue. Seem reasonable?
>
> On Tue, Apr 19, 2016 at 3:50 PM, Lon Varscsak
> wrote:
>
>> Hey guys,
>>
>> I’m
It looks like I had cayenne.DataDomain.sharedCache=true in my model.
Changing this to default solved my issue. Seem reasonable?
On Tue, Apr 19, 2016 at 3:50 PM, Lon Varscsak
wrote:
> Hey guys,
>
> I’m executing a query and then referencing the fetched object and some
> relationships on the fetc
Hey guys,
I’m executing a query and then referencing the fetched object and some
relationships on the fetched object. Later, I create a new ObjectContext
(old one is thrown away) and I issue the fetch again and I notice that the
relationship objects are not refreshed (weirdly I see one of the
rel
Hi Michael,
Although my setup is not exactly the same, there are definite similarities,
namely in the method to fetch a current revision. I may end up reverting the
changes to add the circular FK relationships... it's probably not much
different performance-wise to the server if I have to fetch
Hi Adam,
To follow-up on Andrus' comment, have you considered recursive
relationships? To do so, your TestScript would have getParent() and
getPrevious() type methods/relationships which return a TestScript (if it
exists) plus a parentFK attribute.
I've used the following pattern in the past:
T
I'm super-relieved that this facility exists! Thanks for your attention and
fast reply!
-Adam
From: Andrus Adamchik
Sent: Tuesday, April 19, 2016 3:31 AM
To: user@cayenne.apache.org
Subject: Re: ROP - Define order of inserts on large commit to multiple t
> Is it possible that the new "circular" FK relationships are confusing Cayenne
> into inserting records into the tables in the wrong order?
Yes, it would. Without circular relationships Cayenne provides reliable
operation ordering. With those, the ordering gets messed up.
> If so, is there a
> I would like to have the hashing and validation code on the server to take
> that power away from the client.
Good idea.
> Is it as simple as setting a pre-persist and post-load hook and only
> implementing them on the server?
Yep, callbacks are the way to solve this.
Andrus
> On Apr 16,
Hi Fredrik,
Sorry for late reply. So from the stack trace this is Cayenne 3.1:
https://github.com/apache/cayenne/blob/STABLE-3.1/framework/cayenne-jdk1.5-unpublished/src/main/java/org/apache/cayenne/dba/mysql/MySQLProcedureAction.java#L101
Interesting ... So Cayenne thinks there's a ResultSet, w
11 matches
Mail list logo