IDBroker warning

2004-04-01 Thread T E Schmitz
Hello, I just went through the mailing list archive and there was an unresolved thread back in February regarding the IDBroker warning. "IDBroker is being used with db '', which does not support transactions. IDBroker attempts to use transactions to limit the possibility of duplicate key gene

ant -f build-torque.xml datasql OutOfMemoryError

2004-04-01 Thread Daniel Huang
Hi, I'm trying to generate sql (ant -f build-torque.xml datasql) from a data xml generated by ant -f build-torque datadump. My data xml file is about 1.7M. And I got OutOfMemoryError during the middle of sql generation. I wonder if this is because my data xml is to big or something else. Has anyon

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Jarrell, Maury
> > So it seems that if you use Oracle 9i you can ... > - use Oracle DATE and run with Oracle 8i drivers > - use Oracle DATE, run with Oracle 9i drivers and the Village patch > - use Oracle TIMESTAMP, run with Oracle 9i drivers but don't use dates > in primary keys > > .../Bob Thanks for yo

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Bob Davison
Dave, I am sure you have a good point. I didn't really like it when I came across the 'date in primary key' either but had no real reason to change until now. I guess my mail was to alert anyone else following this route to the problem. We don't all have control over the schemas we have to writ

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Bob Davison
Maury, Many people had problems with Oracle 9i JDBC drivers and Oracle DATE columns, the hours/minutes/seconds were getting set to zero. This was due to a change Oracle made to their drivers when they introduced the TIMESTAMP type in Oracle 9i. What they did was map their DATE datatype to java.s

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Dave Newton
On Thu, 2004-04-01 at 11:02, Jarrell, Maury wrote: > > Totally unrelated, but I thought using "real-world data" as a primary > > key (or portion thereof) was generally a no-no in the DB world. Did I > > misunderstand? > You didn't misunderstand at all. I have had my share of frustrations > dealing

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Jarrell, Maury
> Totally unrelated, but I thought using "real-world data" as a primary > key (or portion thereof) was generally a no-no in the DB world. Did I > misunderstand? > > Dave You didn't misunderstand at all. I have had my share of frustrations dealing with Oracle dates and nulls, and wondered what

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Dave Newton
> From: Bob Davison [mailto:[EMAIL PROTECTED] > Due to the problems with the Oracle 9i JDBC drivers and DATE columns we > moved to TIMESTAMP columns instead. This seemed fine until we used one > of these columns in a primary key. Totally unrelated, but I thought using "real-world data" as a prim

RE: Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Jarrell, Maury
Bob, What were the original problems you encountered? I did a couple of searches on the list archive and couldn't find them. Thanks, Maury -Original Message- From: Bob Davison [mailto:[EMAIL PROTECTED] Sent: Thursday, April 01, 2004 8:59 AM To: [EMAIL PROTECTED] Subject: Oracle 9i TIME

Oracle 9i TIMESTAMP in primary key problems

2004-04-01 Thread Bob Davison
Hi folks, More problems with the Oracle 9i DATE vs TIMESTAMP issue. Due to the problems with the Oracle 9i JDBC drivers and DATE columns we moved to TIMESTAMP columns instead. This seemed fine until we used one of these columns in a primary key. When Torque does an update or delete it builds a