August 6, 2010 at 12:52 pm
My customer's SQL Server 2008 app Selects a timestamp from DB2 on z/OS.
DB2 timestamps on z/OS, converted the Character, are Length 26.
By the time the app sees it, it's length is 23.
So,
2004-11-28-08.01.06.697422 comes back as
2004-11-28-08.01.06.697
Where are we going wrong?
Dave
August 6, 2010 at 11:34 pm
Sql server doesn't support nano seconds as far as I know. You can to use VARCHAR columns to get this info. But as soon as you CONVERT them in date you will lose last three digits...
--lucky
August 9, 2010 at 6:57 am
Maybe we are off track here. Customer is confused (as are we) because this same SQL "works" in Development but not in Production.
The difference we see is this:
In Dev, simple update statement like UPDATE table SET COL02 = 'x' WHERE COL01 = 'y'
gets executed as is.
In Prod, the update statement gets "chagned" to
WHERE COL01 = 'y' AND COL02 = 'm' AND COL03 = 'n' AND COL04 = 'o' AND ....
for every column in the table. (We eventually fail on a date field.)
At first, I figured that there was no unique index or primary key, but the table has all that.
Who is doing the DB2 Catalog query to fetch all columns and changing the Update?
Is there a parameter in SQL Server for that?
Dave
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply