BSharbo (8/3/2015)

---

I can understanding adding an identity key to a table to facilicate joins to other tables.

However in this instance there is only the one table.

I still probably wouldn't make...

Jeff Moden (8/3/2015)

---

ben.brugman (8/3/2015)

---

Thanks for you reply.

I do tend to get hung up to much by 'numbers' in general, and like to...

I've run into the same type of thing over and over (especially, for some reason NOLOCK, it becomes a magical talisman somehow). There are no easy solutions once people get...

A single one-off shrink of the database just is not a big deal. What happens is that people hit issues with disk, transactions, etc., and they set up automated shrinks...

It's just based on how long the pages have stayed in memory. So from one minute to the next, a page has stayed in memory, at least 60 seconds, so...

Whether or not the primary key for the table should be clustered really is down to experimentation and testing. I'm with Tom that, at least from the description provided, you're...

MJ-1115929 (8/1/2015)

---

@Grant Fritchey

Thanks Grant and hope you are doing well, we had chats several times some time back.

I have one query if we want to force the sql server to...

The differences between the servers could be the amount of memory on the servers, the number and speed of the CPU, or just the load because resource contention on any...

Partitioning very rarely helps performance within an OLTP situation. However, if you are looking at a situation where you're rolling off old data, partitioning becomes a pretty viable approach, in...

Ah, it wasn't clear in the first post this is all dynamic SQL.

Try wrapping this in a transaction

exec sp_executesql @s

Brandie Tarvin (7/30/2015)

---

That...

Very true and worth noting.

A pretty hefty percentage of...

Pretty much the same stuff you'd capture if you were monitoring production servers. You want all the wait statistics. You want to see the queues, if any. You definitely should...

The DISTINCT operation is going perform aggregation. You really shouldn't need it if your data structures are correct. That's a pretty big performance hit right there.

SQLRNNR (7/30/2015)

---

Grant Fritchey (7/30/2015)

---

Brandie Tarvin (7/30/2015)

---

As I dig into this issue (Execution Plans HO!), I come across a thread in which one respondent posts a link to Gail's blog and...