Prev: Need help with sequel statement.
Next: install sql server problem
From: Dan on 29 Jul 2010 08:24

"Erland Sommarskog" <esquel(a)sommarskog.se> wrote in message
news:Xns9DC48C25F47C7Yazorman(a)127.0.0.1...
> SnapDive (SnapDive(a)community.nospam) writes:
>> "synatx" was not the right word. Should have used strategy. One thing
>> that came to mind was creating a new varchar(42) column and stuffing
>> the sha1 hash of the column values in there, then I could have a
>> simple unique index on that single column instead of a unique index on
>> the 6 columns. Not sure how that would perform though.
>
> I would go for the index on the six columns, at least at first. You know,
> keep it simple.
>
> Dan suggested CHECKSUM(), but with a few million rows in the destination
> table, you are bound to get rows with the same checksum, even if they
> have different content. A 32-bit value is not enough for that volume.
>

Gah, good point!

--
Dan

First  |  Prev  | 
Pages: 1 2
Prev: Need help with sequel statement.
Next: install sql server problem