Yeah, estimated 1, actual 87000, that's a slight disparity. No wonder it assumed a scan...

Please post at least the execution plan so we can see how the optimizer is resolving the query based on your structures.

The easiest way is going to be to run the RESTORE using WITH MOVE. But, that means the file layout has to be the same, even if the locations are...

It's stored as a clustered index, so there is data stored within the database.

OMG! That explains why I've been doing all this...

Like I said, you probably won't be able to shrink immediately. The unused log space and closed transactions are before the open transactions. Until that changes, you're going to need...

I think the issue must lie somewhere besides the cardinality estimator. In fact, both plans have the exact same estimates. They even have the exact same costs. Something else is...

Like I said, capture the wait statistics that occur during the process. That will tell you what's running slow. You'll have to figure out why.

Look to the row estimates. I'm sure that's probably the issue. The 2014 cardinality estimator is new and if you're seeing differences in execution times, it could be because of...

PHYData DBA (3/16/2015)

---

Grant - I love this article and you are spot on correct.

However I have never read anything that made me think of Star Wars II so much.

You know...

Maybe it's using the IP address. Try that. In general, you won't use the database name to connect from Management Studio. Just the server name. It's also possible you're AD...

The log backup just reclaims the space for committed transactions within the log. It doesn't affect the size of the log at all. If you do need to reduce that...

Dan121 (3/16/2015)

---

Scott, my concern is a size of a row. As Grant said UNIQUEIDENTIFIER is 16 bytes. if I have 503 columns with datatype 'UNIQUEIDENTIFIER' and rows are empty, will...

dan-572483 (3/16/2015)

---

I'm surprised that discussions of whether or not to keep an index seldom include a look at what is generating the query using the index and how important speed...

Without seeing the queries involved, I'm just guessing.

But, it is possible that eve if it's just getting scanned, it's still useful. Let's say your clustered index consists of 1000 pages...