From: Roy P <or...@sh...> - 2007-02-05 16:42:23
|
On Monday 05 February 2007 10:10, Roy P wrote: > On Monday 05 February 2007 09:56, Ashley J Gittins wrote: > > On Tuesday 06 February 2007 00:42, Roy P wrote: > > > When I try create a test dataset in 2.7, it tells me that the id > > > sequence exists. I guess that the id sequence is shared between > > > datasets ? It there an easy way to have the id sequence be unique to > > > each dataset ? Or is there something weird with my database > > > configuration ? > > > > Are you using a separate databases for each? That's the setup that works > > fine for me with two systems. > > Thanks Dieter and Ashley.. > > Hmmm.. ok then it's my config.... > > I'm using separate datasets and SL login, but the same postgresql user. > Are you using a different postgresql user for each one ? > > I've done this before in the past (using the same postgresql user) and > never had an error before.. I must have screwed up somewhere. Looks like i found the problem.... You can tell it's Monday... i'm not fully awake yet... For some reason, the template1 database had a copy of the SQL-Ledger sequences in it. (This must have occured a long time ago on my test box) For those who don't know why this is a problem.... When you create a new database.. the template1 database is automatically used as a... well.... template. Therefore, in this case template1 already contained the sequences and when i tried to create a new dataset it failed because they already existed. Cheers and thanks all. Roy |