Viewing 15 posts - 7,381 through 7,395 (of 7,608 total)
Why not just add a column in the table for the "old" id column?
It's only 4 more bytes, and you'd then be able to reference that column in...
June 13, 2012 at 4:48 pm
Cluster by columnA (date/datetime) only.
That will keep the clustering key short but still restrict the number of rows a query must read and evaluate.
Sometimes additional column(s) are appropriate for the...
June 13, 2012 at 4:40 pm
I wouldn't expect it to change the space available at all, so I'm guessing other activity reduced the space available in the table.
Just to be sure, you did use an...
June 11, 2012 at 12:22 pm
I'm not horribly against the idea of dropping the FK, but maybe I'm not fully understanding the advantage of doing so. How does deleting the parent records first improve performance?
It's...
June 8, 2012 at 2:26 pm
Then why even have the constraint at all then. If all inserts/updates and deletes are done under controlled access methods (ie SPs) and ad-hoc access to the database is not...
June 8, 2012 at 12:23 pm
Might not be a good idea if inserts are expected while also doing the deletes.
Perhaps, if you really normally do INSERTs that would violate the constraint. Hopefully that's extremely...
June 8, 2012 at 9:40 am
I think the point sturner was making was that your order of deletion would violate the foreign key constraint.
That's a good point; yes, I wasn't sure if indeed such a...
June 8, 2012 at 9:28 am
Should I be looking at tuning the fill factor in this situation? Higher or lower than "stock"?
Yes, it can be very applicable to non-clus indexes.
You would lower the fillfactor,...
June 8, 2012 at 9:25 am
That wouldn't work if there were FK from the license_message_value table id --> license_message table.
I reversed the order of the table deletes for performance. By doing the controlling (in...
June 7, 2012 at 8:12 am
If I understand correctly, the license_message table has a:
clus index (/PK) on license_message_id
nonclus index on archive_date
<other nonclus...
June 6, 2012 at 3:44 pm
Here's my version (everybody's got their own version :-)). It provides MB totals for each table, each table and all its indexes, total number of indexes per table and...
June 5, 2012 at 10:43 am
Please try my last posted version.
It includes a check to avoid updating those rows that already have the value that would be assigned.
June 5, 2012 at 9:51 am
@scott: By best practice you never store your unencryption mechanisms and your data in the same place. Admittedly, there are times when you may need to work with the data...
June 4, 2012 at 5:06 pm
I suggest looking into SQL's existing EncryptByKey and DecryptByKey functions.
I disagree with the idea that encryption/decryption can be left to the app. Because you may want to work with...
June 4, 2012 at 4:35 pm
Maybe I'm not getting this, but tableB does not get updated.
No, but tableB provides a lookup table to get the value with which to update the other table. ...
June 4, 2012 at 3:25 pm
Viewing 15 posts - 7,381 through 7,395 (of 7,608 total)