I suspect the answer may be in your I/O subsystem, as various people here have commented that the CPU is not over-stretched and you are reported a 0 CPU Queue...

I'd agree with other posters that 8 drives is limiting.

Without having experienced RAID 6 I can't really comment on that.

I would go for a mixed setup.

Take 2 drives in RAID...

I'd create a new table and name it something significantly different.

I've done it the other way occasionally (same table and a flag to indicate which data type it is included...

This is the problem with this type of Primary Key.

I would say yes, you should create an index (or unique constraint) on the data that you expect to be unique....

My experience with this kind of thing is that allowing RAID to handle the load balancing across multiple spindles is USUALLY better than trying to manually allocate the data extents...

If there can only be 2 rows, then probably the simplest is to self join table1

So your query ends up something like:

SELECT

t1.leg,

t1.ccy

...

What are the rules for this? Is it based on a column in the data being the same?

I'm not sure if this is relevant to this question, but one note I will make is that to insert a repeating XML structure inside another XML Structure you must...

The entire view may need to be instantiated in tempdb.

Try changing to the underlying table.

Alternatively, you could try an indexed view

If you just need to add records to A, try something along the lines of

INSERT A

SELECT B.*

FROM B

LEFT OUTER JOIN A

WHERE...

Great to see this technique reviewed again. I came across it in Ken Henderson's book "The Guru's Guide to Transact-SQL" and have used it time and time again to normalise...

Silly little thing, but we use a case sensitive collation.

Change "cte" in the actual select near the bottom to be "CTE"

I may be a bit strange in thinking this, but to my mind a DBA should understand what is being deployed (ultimately because we are the ones to support it!)...

This is an interesting article, which left me scratching my (SQL 2000) head over the syntax. Once I had that straight and looked up ROW_NUMBER in the Books Online I...

Checkpointing is a process where all the "dirty pages" (changed data) are written to the disk. It is described fairly well in the SQL help.

The main cause of issues around...