I was writing a script to import a 3,5M records table. After about 800k records the system gives a memory error (I monitored the memory free decrease and it happens always at the same "level").
If I limit the SQL query at 800k and start again with the following 800k I can do it all.
When I save the file takes much much less.
Anybody can point me to some reading to explain why this happens? Nothing too difficult ... 😉
Obviously if I go on my other computer with more memory and Vista64 the limits will change but not the "process" that I'd like to understand.
On 64 bit Vista the behavior is different. I admit that there is more memory (4Gb)... it looks like it gets a lot of buffer space but uses very little of memory and arrives to the end in one single shot.
