mitzyturbo (11/1/2016)

---

As I have 2 log tables to pull data from, how would this work?

do you mean by just having a clustered index by the modified date on both log...

Properly clustering should still allow a merge join, which should be much faster than anything you're having to do now.

Cluster the log table on the relevant arrival datetime that you are comparing by (rather than, presumably, a meaningless identity value). Then query directly from the table, joining to...

drew.allen (10/31/2016)

---

Latin is the base for ALL Romance languages. I have an MA in Linguistics, so I have some idea what...

No, actually COBOL had numeric values as well. For example, packed decimal...

CELKO (10/31/2016)

---

Yep, and nobody takes Latin either. Yet everywhere you...

david.jack (10/31/2016)

---

ScottPletcher (10/28/2016)

---

david.jack (10/28/2016)

---

VAL1 can return NULL unfortunately so using a...

I doubt the vast majority of developers today have even heard of COBOL, and they certainly have never programmed in it.

Yes, it can,...

CELKO (10/28/2016)

---

You do not understand the difference between a string and a numeric in SQL. Registration numbers are called "tag numbers" in data modeling; they are a name of an...

david.jack (10/28/2016)

---

VAL1 can return NULL unfortunately so using a comparison operator...

Seems more confusing than the natural CASE statement to me:

SUM(CASE WHEN VAL1 <> 0 THEN VAL1 ELSE VAL2 END)

You're right. Nvarchar shouldn't be used unless it's needed or realistically could be needed.

For example, if you create a two-character code column, there's no reason to use nvarchar, since...

For best performance, avoid the separate EXISTS lookup as well:

SELECT O.ow_id

FROM props.dp_owner O

INNER JOIN props.dp_entry E

...

Just off the top of my head, I'd do something like below ... if no backup date appears, then there was no full backup after the last restore.

SELECT tdb.*, rh.restore_date...