D'OH!

Here's my script, based on Books Online's description of how to calc max row size. I think it will give you a reasonably good estimate of the max possible...

Yep. To me, the RAM restrictions and the lack of data compression are the big factors in Standard Edition. Snapshots not being available isn't that bad, although it...

Just let SQL tell you :-).

Create an empty table and then run this command on it:

SELECT *

FROM sys.dm_db_index_physical_stats(DB_ID(),OBJECT_ID('<your_table_name>'),NULL,NULL,'DETAILED')

It will give you the min and max row sizes.

djj (2/27/2015)

---

The only unique things in the table are IDX and GUID. Using a combination of columns it is still not possible to get a unique combination. Since...

Eric M Russell (2/27/2015)

---

A non-clustered index won't necessarily result in double I/O on reads. Given a well designed non-clustered index with all required columns included, it can entirely cover the...

Before making such a critical choice, review SQL's missing index and index usage stats. Odds are, you have a better choice than an identity column for the clustering index...

Eric M Russell (2/27/2015)

---

N_Muller (2/26/2015)

---

...

...

To note is the fact that the table(s) will have millions of rows, but the customer will request data for at at most, 100 or so...

Eric M Russell (2/27/2015)

---

ScottPletcher (2/27/2015)

---

A lower FILLFACTOR might be required. Since users always provide the up-to-50-byte key, the idea is to avoid having to create nonclustered indexes. This...

Jeff Moden (2/27/2015)

---

ScottPletcher (2/27/2015)

---

Eric M Russell (2/27/2015)

---

As a general rule, the clustered key should be sequentially incremented (to avoid page splits and accumulating fragmentation) and it should be narrow, because...

Eric M Russell (2/27/2015)

---

As a general rule, the clustered key should be sequentially incremented (to avoid page splits and accumulating fragmentation) and it should be narrow, because it will be...

Lynn Pettis (2/27/2015)

---

ScottPletcher (2/27/2015)

---

Lynn Pettis (2/26/2015)

---

ScottPletcher (2/26/2015)

---

Either cluster directly on the string itself. Or, if you don't want to do, encode the string into an int (or smallint if...

Lynn Pettis (2/26/2015)

---

ScottPletcher (2/26/2015)

---

Either cluster directly on the string itself. Or, if you don't want to do, encode the string into an int (or smallint if you have less...

Depending on the specifics, it might be less overhead to save what you need from the summary tables, truncate them, then do a bulk insert of everything back to them...

Change to money rather than more decimal places. I believe the law now restricts to 4 decimal places when calculating most interest accumulation.

Also, let's look at SQL's missing index and index usage stats for the table:

USE [<your_db_name_here>] --change to desired db if not already there

SET DEADLOCK_PRIORITY LOW --probably irrelevant, but just in...