What's really critical here is how the 680MrowsTable is clustered. Best for what you are doing here would be ( CustomerId, ACTdatetime ). If this is typical of...

If UAN is how you do lookups and, especially, joins, then you'd probably get better overall performance by clustering the table on UAN rather than on an identity column. ...

Or:

id bigint IDENTITY(12345000000000, 1) NOT NULL,

...

I'd be more concerned with log waits and/or dynamic log growth than with log flushes per second. If you're writing a lot of data, you will have a lot...

This code may give you some ideas. It looks at the most recent <nn> identical activities and averages them to get an estimate:

declare @num_of_most_recent_activities_to_use_to_estimate int

set @num_of_most_recent_activities_to_use_to_estimate = 5

create table...

Typically I've found that DISTINCT is faster than GROUP BY when SELECTing a single column, i.e. as below, but obviously that could vary:

SELECT DISTINCT

...

There's extremely little detail here, but it seems as if you are not distinguishing between a course and a single offering of the course, which I'll call a "class" for...

gary1 (1/13/2012)

---

Thanks azdzn.

In Oracle, from toad for Oralce, if we select a table and see the Script, it shows complte table script along with the grants

But in SSMS, I can...

Jeff Moden (6/16/2016)

---

ScottPletcher (6/16/2016)

---

Jeff Moden (6/15/2016)

---

Sergiy (6/15/2016)

---

Actually, your "simple question" does not have a correct answer.

The only correct answer would be a question in response.

Probably, more that one question in...

Because sys.dm_db_index_physical_stats is so much overhead, you'll want to filter out non-heaps before you do the CROSS APPLY to that function.

Jeff Moden (6/15/2016)

---

Sergiy (6/15/2016)

---

Actually, your "simple question" does not have a correct answer.

The only correct answer would be a question in response.

Probably, more that one question in response.

Until your requirements...

I agree, definitely save, truncate and (re)load.

Btw, you should very likely change the clustered index on the table to insert datetime if it's currently based on an identity column. ...

I'd use separate dbs rather than separate schemas. We are client-based as well, and some of our clients demand that their data is isolated from the data of all...

Eric M Russell (6/15/2016)

---

Jeff Moden (6/14/2016)

---

Eric M Russell (6/14/2016)

---

You'll rarely ever be asked pop quiz style questions like "What is a clustered index?" or "How do you get the...

It's my understanding -- although I can't construct code right now that confirms it -- that:

SET LOCK_TIMEOUT

can cause that type of behavior. I had faith in that claim because...