Re: Confirming \timing output - Mailing list pgsql-general

From Steven Schlansker
Subject Re: Confirming \timing output
Date
Msg-id CEA07360-490A-4578-8FEB-03577FC73C66@likeness.com
Whole thread Raw
In response to Confirming \timing output  ("Gauthier, Dave" <dave.gauthier@intel.com>)
List pgsql-general
On Aug 23, 2012, at 11:13 AM, "Gauthier, Dave" <dave.gauthier@intel.com> wrote:

> With \timing set on, I run an update statement and it reports....
>
> Time: 0.524 ms
>
> Is that really 0.524 ms?  As in 524 nanoseconds?

0.524ms = 524000ns

Perhaps you meant microseconds?

0.524ms = 524us

If all your data happens to be in RAM cache, simple queries can execute very fast!  Unless you have a reason to believe
it'swrong, I would trust it to be accurate :-) 

>
> Also, is this wallclock time or some sort of indication of how much cpu it took?
>
> Thanks for any answers !
>


\timing measures wall time.  There's a more detailed discussion of the difference between this and e.g. EXPLAIN ANALYZE
here:

http://postgresql.1045698.n5.nabble.com/What-does-timing-measure-td4289329.html





pgsql-general by date:

Previous
From: Sébastien Lorion
Date:
Subject: Re: Amazon High I/O instances
Next
From: John R Pierce
Date:
Subject: Re: Confirming \timing output