Hi,

Just a little bit confused. When create a clustered columnstore index, if all columns will be included, is that the same as rowstore index? What's the benefit?

Thanks

None of the options is 100% accurate. 🙂

For the last one, it's counting NULL. I think that's why it returns 0.

Good explanation, thanks.

Don't think 'nonclustered index' shoud be included in the question.

If you have to keep the data this way, maybe can parse the hourse_no, then order by different parts.

SELECT

CAST(SUBSTRING(hourse_no,1,CHARINDEX('-',hourse_no)-1) AS INT),

CAST(SUBSTRING(hourse_no,CHARINDEX('-',hourse_no)+1,2) AS INT),

...

FROM #temp

ORDER BY

CAST(SUBSTRING(hourse_no,1,CHARINDEX('-',hourse_no)-1) AS INT),

CAST(SUBSTRING(hourse_no,CHARINDEX('-',hourse_no)+1,2) AS INT),

...

good question.

Very good question and good explanation.

nice function.

tricky, tricky.