Hi all,


today I've found an interesting "feature" while trying to generate lots
of dummy data rows.

I've created a small table with a ID and NAME column (ID is an
autoincrement column) and inserted one row.

My intent was to generate 1, 2, 4, 8, 16... records via

INSERT INTO test(name) SELECT name FROM test;

but the server runs and runs and runs.. and the database file grows up
to ~2gb.

With INSERT INTO test(name) SELECT FIRST 1000000 name FROM test;
I was able to produce one million records without any script - really
nice.

But I think this is not really a feature.
Running 10-50 simultanous statements like this one, the hard disks
explodes and the server goes down :-)


Should I write a bug report or is this an intended feature?

(tested with Win32 FB SS 2.0.5)


Daniel




--
web: http://daniel.vogelbacher.name
irc: cytrinox @ (freenode|ircnet|quakenet)