Re: : Tracking Full Table Scans

[Venkat Balaji <venkat.balaji@xxxxxxxx>]

Thanks Kevin !!

I will have a look at the source tree.

Regards

VB

On Tue, Sep 27, 2011 at 10:45 PM, Kevin Grittner <Kevin.Grittner@xxxxxxxxxxxx> wrote:

Venkat Balaji <venkat.balaji@xxxxxxxx> wrote:

> I would like to know the difference between "n_tup_upd" and
> "n_tup_hot_upd".

A HOT update is used when none of the updated columns are used in an
index and there is room for the new tuple (version of the row) on
the same page as the old tuple.  This is faster for a number of
reasons, and cleanup of the old tuple is a little different.

If you want the gory implementation details, take a look at this
file in the source tree:

src/backend/access/heap/README.HOT

-Kevin