Posted by nnelson on 02/25/06 07:09

I have a base customer table of 2 million records. We're doing some
testing and I added 33000 rows incorrectly. No biggie, we'll just
delete them, right? Nope....9 hours later, process is still running
using this query.

delete from customer where custid in (select custid from #tmp1)

k...so we'll change it.

delete from customer where custid > 2295885

2 hours later, still going slower than a turtle...literally 1 row per
second.

custid is my primary key so you can't tell me it's not indexed. It's
pretty bad when I have to resort to doing a restore of my entire
database of 27GB because it's faster than deleting 33000 rows.

WTF?

[Back to original message]