Feedback-ID: i76984098:Fastmail
Message-ID: <5b89abb9-1c71-4836-9cf6-56c694e80f01@aklaver.com>
Date: Sun, 18 May 2025 12:29:05 -0700
MIME-Version: 1.0
User-Agent: Mozilla Thunderbird
Subject: Re: pg_stat_statements has duplicate entries for the same query &
 queryId
To: Jevon Cowell <jcowell@atlassian.com>, pgsql-general@lists.postgresql.org
References: <CAMAx+cJM2Z32xvVbduGsF7rgOL6HuRcqq2Pob6w33wS1q8U0hw@mail.gmail.com>
Content-Language: en-US
From: Adrian Klaver <adrian.klaver@aklaver.com>
In-Reply-To: <CAMAx+cJM2Z32xvVbduGsF7rgOL6HuRcqq2Pob6w33wS1q8U0hw@mail.gmail.com>
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Archived-At: <https://www.postgresql.org/message-id/5b89abb9-1c71-4836-9cf6-56c694e80f01%40aklaver.com>
Precedence: bulk

On 5/18/25 12:20, Jevon Cowell wrote:
> Hi Folks!
> Let me know if there's a better mailing list to ask this in.
> 
> I have a statistics collector that collects data from various postgres 
> statistics tables, pg_stat_statements being one of them. This is done on 
> an entire fleet of Postgres databases. From the collected data we record 
> the timestamp of each collection in the query itself as extract(epoch 
> from now()) as ts. What I'm seeing is that for the same query 
> /and/ query id, there are two rows with different statistics data /at 
> the same time/. For example one row can have 2 calls while another can 
> have 4. Anyone else run into this or have any idea why this can occur?

 From here:

https://www.postgresql.org/docs/current/pgstatstatements.html

queryid bigint

Hash code to identify identical normalized queries.

query text

Text of a representative statement


 From here:

https://www.postgresql.org/docs/current/functions-datetime.html#FUNCTIONS-DATETIME-CURRENT

"now() is a traditional PostgreSQL equivalent to transaction_timestamp()"

and

"transaction_timestamp() is equivalent to CURRENT_TIMESTAMP, but is 
named to clearly reflect what it returns."

Therefore now() is pinned to the time the transaction started. 
Consequently it is conceivable that the queries actually ran at 
different times but got stamped with an identical timestamp via 
extract(epoch from now()) as ts.

> 
> Regards,
> Jevon Cowell

-- 
Adrian Klaver
adrian.klaver@aklaver.com