ensure_unique_url very slow and memory consuming on large datasets about stringex HOT 6 CLOSED

already planning on using batched finds here. open to other ideas as well.

from stringex.

is would use SQL count on attr_name with "SOME%" this would give a good starting point,
after that I would generate the next say 100 possible results, try to find these all at once using SQL IN (possible_20, possible_21, ...) and then try to find the first not existing result with ruby, this you can repeat without to worry about preformance.

from stringex.

got a patch idea?

from stringex.

master uses find_each to go through smaller sets. short of a patch submission i don't have a clear idea here. feel free to submit a pull request or whatever. happy thanksgiving as well.

from stringex.

@rsl we ran into another similar issue. Expect a patch shortly.

from stringex.

closing this as it seems dead. reopen if needed.

from stringex.