Posted by ja on 10/01/21 11:39

Hello,

I have a table that contains approx 2 million records. I want to
query 2,000 random records.

Example: 500,000 names divided by 20,000 samples = 25. The N is 25,
meaning samples would be sent to every 25th name on the list.

this is what I am using to create a random selection

order by substing(f1,4,1) + substing(f2,3,1) + substing(f3,3,1) +
substing(f4,2,1)

This method takes a very long time.

Any suggestions????


TIA


ja

[Back to original message]