Re: parse/bind/execute - Mailing list pgsql-performance

From David Johnston
Subject Re: parse/bind/execute
Date
Msg-id CAKFQuwYD1-n5V=yKCDajQSVsuC2MeOzKXXPy1N4N2x0dm41yTA@mail.gmail.com
Whole thread Raw
In response to Re: parse/bind/execute  ("Huang, Suya" <Suya.Huang@au.experian.com>)
List pgsql-performance
Please do not top-posts on these lists.

On Wednesday, June 4, 2014, Huang, Suya <Suya.Huang@au.experian.com> wrote:
Thank you David, I copied the detailed activity from the report as below. As it shows, it has prepare and bind queries.  One of the item has Bind/Prepare pretty high as 439.50. so that looks like a good value?

Another question is if bind only happens in a prepared statement?

Day     Hour    Prepare         Bind    Bind/Prepare    Percentage of prepare
Jun 03  00      205     209     1.02    1.27%
        01      19      19      1.00    0.17%
        02      0       0       0.00    0.00%
        03      0       0       0.00    0.00%
        04      6       6       1.00    0.00%
        05      2       879     439.50  0.02%
        06      839     1,323   1.58    7.01%
        07      0       0       0.00    0.00%



Yes. Something that high usual involves batch inserting into a table.  To be honest, a global picture is of limited value for this very reason.  Representing all of your usage as a single number is problematic.  Breaking it down by hour as done here increases the likelihood of seeing something useful but typically that would be by chance.  In this case because batch processing is done in the early morning and few users are probably on the system (a common source of one-off statements) the numbers here are dominated by the special case of bulk inserts and are not typical of normal activity and performance.

David J.



  

pgsql-performance by date:

Previous
From: "Huang, Suya"
Date:
Subject: Re: parse/bind/execute
Next
From: Vince Lasmarias
Date:
Subject: High CPU load when 'free -m' shows low 'free' memory even though large 'cached' memory still available