On Fri, 02 Jan 2004 16:21:06 +0100, Armin Waibel wrote: > Hi, > > Gunnar Hilling wrote: > >> Hello, >> >> what can I do to prevent "Out of Memory"-Errors when working on large >> datasets? >> >> I tried to do "tx.commit(); tx.begin();" after doing part of the (big) >> job, but I didn't succeed. >> Do I have to empty the Cache explicitly? >> What is recommended? >> I create the objects that cause the error in my application, so I thought >> they should be garbage-collected... or aren't they? >> > hmm, if implicit locking is set true, all queried objects will be read > locked. Read lock based on object itself, thus GC can't remove objects > with read-lock. > Try to set implicit locking to false when query large datasets. Or you > can try to use paging from the PB-api (don't know if odmg-api supports > paging). > using PB-api doesn't help. I tried to set implicit locking to false, but get errors about "dlist_id null values" (see thread "strange error" from last night) I think when updating values I would have to write lock some objects myself but how can it tell which ones?
thanks, -Gunnar --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
