>But these special optimizations only apply when min(), max(), and
count(*) are used in isolation. Hence, they do not work for the first
query above that uses all three functions at one.
Thanks Richard.
>(1) If you are using INTEGER PRIMARY KEY, you should *not* be using a
WITHOUT ROWID. You
On 8 Dec 2017, at 1:17pm, Simon Slavin wrote:
> helps you (and us, if you want to post it) to understand what’s happening.
Please ignore my post. Dr H explain your situation exactly.
Simon.
___
sqlite-users mailing list
On 8 Dec 2017, at 12:20pm, x wrote:
> I have a table with 2.4 million records. It’s a without rowid table (I don’t
> know if that’s significant) with an integer primary key (ID) and several
> secondary indexes of the form (OtherCol, ID). If I run
>
> select min(ID),
On 12/8/17, x wrote:
>
> I have a table with 2.4 million records. It’s a without rowid table (I don’t
> know if that’s significant) with an integer primary key (ID) and several
> secondary indexes of the form (OtherCol, ID). If I run
(1) If you are using INTEGER PRIMARY
I have a table with 2.4 million records. It’s a without rowid table (I don’t
know if that’s significant) with an integer primary key (ID) and several
secondary indexes of the form (OtherCol, ID). If I run
select min(ID), max(ID), count(*) from BigTbl;
It takes 0.67 secs
If I run the three
5 matches
Mail list logo