Re: SQL Tools and Performance

On Thu, Mar 12, 2015 at 12:38 PM, <rob@xxxxxxxxx> wrote:

<snip>
Also, 600,000 records might simply be too much data. Sending all that
across the network could be what's slow.
</snip>

"Also" is the thing that is wrong. That IS what I am saying. Not only
are you sending all those records down to the client so it can then decide
which rows really matter, but you're making the SQL engine on that client
do that work. You should really find out a way to get the 'where' to
execute on the server. See my previous suggestions.

But OP said right up front that the result set *IS* about 600,000
records! In other words, after WHERE is already done filtering,
what's LEFT is 600,000 records.

Go ahead and try generating an Excel file with that amount of data.
You can do it with a .xlsx, but... it's not fast.

And I *still* don't understand what you mean by "the SQL engine on
that client". My contention all along is that the client doesn't have
an SQL engine. Why would it? The data lives on the server. The
server also has its own (typically quite powerful) SQL engine. Why
would the client have its own?

John Y.