Date: Sun, 16 Apr 2023 12:15:11 -0500
From: Justin Pryzby <pryzby@telsasoft.com>
To: Clemens Eisserer <linuxhippy@gmail.com>
Cc: pgsql-performance@postgresql.org
Subject: Re: speeding up grafana sensor-data query on raspberry pi 3
Message-ID: <ZDwtHyMicQ3+6AxE@telsasoft.com>
References: 
 <CAFvQSYR_kaOm8xQjzXUWiUSvYmEXiShvDBimVNZTUtN-3W3TLg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: 
 <CAFvQSYR_kaOm8xQjzXUWiUSvYmEXiShvDBimVNZTUtN-3W3TLg@mail.gmail.com>
Archived-At: 
 <https://www.postgresql.org/message-id/ZDwtHyMicQ3%2B6AxE%40telsasoft.com>
Precedence: bulk

On Sun, Apr 16, 2023 at 07:00:33PM +0200, Clemens Eisserer wrote:
> Hi,
> 
> I am currently trying to migrate an influxdb 1.7 smarthome database to
> postgresql (13.9) running on my raspberry 3.
> It works quite well, but for the queries executed by grafana I get a
> bit highter execution times than I'd hoped for.

Suggestions:

 - enable track_io_timing and show explain (analyze,buffers,settings)
 - or otherwise show your non-default settings;
 - show \d of your table(s)
 - show the query plan for the 6 months query .  The query plan may be
   different, or (if you can run it with "analyze") it may be
   illuminating to see how the query "scales".
 - consider trying postgres 15 (btw, v16 will have a beta release next
   month)

https://wiki.postgresql.org/wiki/Slow_Query_Questions

-- 
Justin