What kind of datatype should one use to store hashes?
Solution 1:
You should use the binary
datatype. You can use binary
instead of varbinary
because the hash function will always return the same number of bytes for the same type of hash (e.g. MD5
, SHA1
, etc.). This will cut down on the (slight) overhead required to manage a variable length binary (varbinary
) column.
In terms of what size to make it, you can run this query to check the length of each hash type:
SELECT DATALENGTH(HASHBYTES('MD2', 'Testing')) AS [MD2Length],
DATALENGTH(HASHBYTES('MD4', 'Testing')) AS [MD4Length],
DATALENGTH(HASHBYTES('MD5', 'Testing')) AS [MD5Length],
DATALENGTH(HASHBYTES('SHA', 'Testing')) AS [SHALength],
DATALENGTH(HASHBYTES('SHA1', 'Testing')) AS [SHA1Length],
/* 2012 only: */
DATALENGTH(HASHBYTES('SHA2_256', 'Testing')) AS [SHA2_256Length],
DATALENGTH(HASHBYTES('SHA2_512', 'Testing')) AS [SHA2_512Length];
And it should come out with this:
MD2Length MD4Length MD5Length SHALength SHA1Length SHA2_256Length SHA2_512Length
--------- --------- --------- --------- ---------- -------------- --------------
16 16 16 20 20 32 64