Thanks, my stupid mistake. It really helps. And any reading, could u suggest, please, for heavy database computing, any tips, I really need it, since I am building a microsimulation model contains millions record, for consistency, I am running for a database approach as a cache, otherwise, it might be ok for 2 million, but out of memory for 20 million, etc.
Thanks Jin -----Original Message----- From: Mamta Satoor [mailto:[EMAIL PROTECTED] Sent: 18 September 2004 17:04 To: Derby Discussion Subject: Re: Speed for millions updates Hi Jin, One thing that comes to my mind is maybe you are running your updates in Derby's default mode of autocommit true. Putting the updates in a transaction by setting autocommit to false on the connection will definitely speed things up in case you are not doing it. Mamta Jianhui Jin wrote: > Hi guys > > My database is about 60 million records, has simple data types, int or > float. My issue is the speed to update them. My model will run on > those records and update them in one run. > It seems it require huge amount of time to update them, which is not > practical for me. > > How to make it fast. I tried using batch, didn't help, not much > difference to using PreparedStatement. > > Any helps. > > Jin
