From: | Erwin Brandstetter <brsaweda(at)gmail(dot)com> |
---|---|
To: | pgsql-general(at)postgresql(dot)org |
Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
Subject: | Re: timestamp(0) vs. timestamp |
Date: | 2011-04-28 12:33:55 |
Message-ID: | 4DB95EB3.80503@gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
On 27.04.2011 19:36, Tom Lane wrote:
> Erwin Brandstetter<brsaweda(at)gmail(dot)com> writes:
>> Hi all!
>> This may seem unimportant, but I still would like to know.
>> I have columns for timestamps without fractional digits, so I could
>> define them as timestamp(0).
>> However, there is no way fractions could ever enter anyway, because
>> triggers and / or checks guarantee values without fractional seconds.
>> Is it advantageous to define the column as timestamp(0) or simply as
>> timestamp?
>> Does the query planner or any other part of the RDBMS profit from the
>> additional information in the definition?
>> Or do I just create an overhead of useless checks or conversions
>> (however small)?
> There's no advantage to the system from knowing that, but consider that
> having timestamp_in force the values to be nonfractional is likely to be
> faster than having logic in a trigger to do it.
>
> regards, tom lane
Thank you for the insight.
Is there any kind of overhead for timestamp_in due to the precision modifier - if non-fractional values are given?
Regards
Erwin Brandstetter
From | Date | Subject | |
---|---|---|---|
Next Message | Sim Zacks | 2011-04-28 12:49:14 | Re: plpython module import errors |
Previous Message | Oleg Bartunov | 2011-04-28 11:40:15 | Re: GIN index not used |