Date Time conversion/ Date diff in SSRS - sql

Question

I have an SSRS report where I am using below query.
(This query works fine in SQL server, problem is only in SSRS report)
--DECLARE #Range Number = 10;
SELECT * FROM TBL1 WHERE
USERNAME = 'MIKE'
AND
(
#Range = '10'
and
Convert(datetime, MyDate, 120) <= GETDATE()
)
or
(
#Range IN ('20','30')
and
DATEDIFF(DD, Convert(datetime, MyDate, 120) , GETDATE()) <= #Range
)
Unfortunately the myDate Column coming from database is a varchar column.
The SSRS throws an out-of-range exception.
Then I tried converting Getdate to Convert(varchar(10), getdate(), 120) and compare with myDate (without conversion as myDate is already in YYYY-MM-DD format but as a varchar in database)it still throws error. I assume this time coz SSRS is not able to process datediff in varchar columns.
When I run these queries individually, it works fine. i.e -
Declare #Range Number = 10;
Select * from tbl1
where username = 'MIKE' and
(
#Range = '10'
and
convert(datetime, MyDate, 120) <= getdate()
Has anyone faced similar issue in SSRS ???

Answer 1

Try this!
SELECT * FROM TBL1 WHERE
USERNAME = 'MIKE'
AND
(
#Range = '10'
and
cast(myDate as datetime) <= GETDATE()
)
or
(
#Range IN ('20','30')
and
cast(myDate as datetime) , GETDATE()) <= #Range
)

Answer 2

Presumably, your problem is that some values of myDate are not in the correct format. If you are using a more recent version of SQL Server, the function try_convert() can be a big help.
The reason it is failing is because of your logic. The second condition after the or is looking at all records, not just Mike's. I think you intend for the where clause to be:
SELECT *
FROM TBL1 WHERE
WHERE USERNAME = 'MIKE' AND
((#Range = '10' and Convert(datetime, MyDate, 120) <= GETDATE()
) or
(#Range IN ('20','30') and
DATEDIFF(DD, Convert(datetime, MyDate, 120) , GETDATE()) <= #Range
)
);
Note the extra set of parentheses. Now, this will probably help for this particular query for MIKE. But it won't help overall. Finding the value(s) that fail conversion can be daunting. If you are lucky, they fail easily. You can look for them with starting with:
select MyDate
from tbl1
where MyDate not like '[0-9][0-9][0-9][0-9]-[0-2][0-9]-[0-3][0-9]%';
If you are lucky, then this will find the offending values. Otherwise, you'll have to dive more deeply into the date formats, looking at the particular month and day (and potentially hour, minute, and second) values.
Moral: Store dates as dates. Don't store them as varchar().