declare @cursor cursor, @userID int
set @cursor = cursor for
(select id
from users)
open @cursor

while 1=1
begin
fetch from @cursor into @userID;
if @@fetch_status 0
break

update users set reportMonthDay = 1 + CONVERT(INT, 28*RAND())
where id=@userID

end

Great Tip! thanks!

I dont want just the first top 100, I want 100 random records and I am not trying to generate a new GUID. Which is why I use it in the ORDER BY clause. It works good for my purposes!

Say there are 5 records with PostalCode=98011. The reader may want to retrieve any one of these 5 records when running a query. Next time the query is run, he may want to retrieve another random record with same PostalCode

?

It maybe, he’s exploiting the random number for session management or presenting the result in some format in front-end/reports or working with some existing or new design pattern that requires it.

select
rand(PostalCode) as RandomNumber, PostalCode
from
Person.Address
where
StateProvinceID = 79
order by
PostalCode

The RAND() function will only return one unique value for each seed, so if you pass it a common seed all of those rows would share the same random value.