Most likely, what you need to see is what DML commands are being run at the time the database grows. To do that, you can either run a trace,...

He said "behave", he didn't say how to behave ...

Actually, you can, but indirectly. Put an "Instead of Update" trigger on the view, and have the trigger parse out the update into the three tables.

Warning, this can cause...

david.ashby (11/16/2012)

---

Cost for cheap SAN storage where I work is £3,000 per TB ($4,500 USD) and £10,000 per TB ($16,000 USD) for the 5 year TCO. As you can imagine...

Total cost of storage is a complex evaluation. Trying to calculate return-on-investment with it is virtually impossible, too. Far too many variables.

Despite that, I know what the dollar...

I'm going to suggest taking each of the subqueries, from the inside first, and turning it into a temp table, then using the final temp table in the final query.

The...

Index defragmentation shouldn't cause a performance degradation, but it won't necessarily speed things up either. It depends on how the index is being used.

Single-row seeks, for example, aren't significantly...

Jeff Moden (11/16/2012)

---

@aaron,

Great explanation and write-up. Thanks for takinng the time.

The idea of storing XML is definitely a good one for this type of thing. Since I haven't...

EAV vs XML is a complex subject. I prefer XML because it can be treated as an atomic item in a single row of a table.

Here's an example of...

Since I was out a few days, I missed this article till today.

Interesting methodology.

Testing it on my workstation, I got 45 seconds for a million-row hierarchy, 30 nodes deep.

On 10k...

Because I use custom maintenance scripts, I maintain a log of "actions done per database". Rebuild some indexes? Log it. Run a backup? Log it. ...

Yeah. SSIS has some serious holes in it, especially when it comes to forcing specific formats on data that doesn't match the collation of the server. It's like...

What you have there is a spreadsheet, not a database table. Rows in a database table don't have a sequence to them, so there's no way for the database...

Count() just gets the number of rows, by default.

Here's a sample of how it works. Run this in a test environment so you can see what I'm talking about:

IF...

You might try "count (distinct a.id)" and see if that does what you need. That will count distinct values, instead of total rows.

But it's normal for a join from...