Re: BUG #15678: The postgresql.conf string does not define server behavior (default_transaction_isolation)

The following bug has been logged on the website:

Bug reference: 15678
Logged by: Vasiliy Korotkikh
Email address: dhaenoor(at)yandex(dot)ru
PostgreSQL version: 11.2
Operating system: Debian Stretch
Description:

In the configuration file of my server there is the following line:
default_transaction_isolation = 'serializable'

I make a query like this:
select current_setting('transaction_isolation')

It displays this message:
read committed

The isolation level inside the functions also does not correspond to the one
specified in the configuration file.
For verification, I add into functions something like this: INSERT INTO
t1(textval) VALUES (current_setting('transaction_isolation'));

Also I used next query: ALTER ROLE ALL IN DATABASE mydb SET
default_transaction_isolation TO 'serializable';

P.S. If in the future you want to get more sensible error reports, then you
should think about a few examples. People understand examples better than
lengthy documentation descriptions.

On 2019-03-08 17:38:33 +0000, PG Bug reporting form wrote:
> The following bug has been logged on the website:
>
> Bug reference: 15678
> Logged by: Vasiliy Korotkikh
> Email address: dhaenoor(at)yandex(dot)ru
> PostgreSQL version: 11.2
> Operating system: Debian Stretch
> Description:
>
> In the configuration file of my server there is the following line:
> default_transaction_isolation = 'serializable'

Did you reload the configuration after making that change? SELECT
pg_reload_conf();

Greetings,

Andres Freund

<div xmlns="http://www.w3.org/1999/xhtml"><div>Problem solved.</div><div>The reason was that I used pgcli.</div><div>When I use psql, everything works as it should.</div><div>I assume that pgcli chooses transaction isolation level by itself.</div></div><div xmlns="http://www.w3.org/1999/xhtml"> </div><div xmlns="http://www.w3.org/1999/xhtml"><a href="https://www.sql.ru/forum/1310005/funkciya-v-pg11-2-ne-zapuskaetsya-s-urovnem-izolyacii-serializable">https://www.sql.ru/forum/1310005/funkciya-v-pg11-2-ne-zapuskaetsya-s-urovnem-izolyacii-serializable</a></div><div xmlns="http://www.w3.org/1999/xhtml"> </div><div xmlns="http://www.w3.org/1999/xhtml">Thank you!</div><div><br /></div><div><br /></div><div>08.03.2019, 22:13, "Andres Freund" &lt;andres(at)anarazel(dot)de&gt;:</div><blockquote type="cite"><p>On <span>2019-03-08 17</span>:38:33 +0000, PG Bug reporting form wrote:<br /></p><blockquote> The following bug has been logged on the website:<br /><br /> Bug reference: 15678<br /> Logged by: Vasiliy Korotkikh<br /> Email address: <a href="mailto:dhaenoor(at)yandex(dot)ru">dhaenoor(at)yandex(dot)ru</a><br /> PostgreSQL version: 11.2<br /> Operating system: Debian Stretch<br /> Description: <br /><br /> In the configuration file of my server there is the following line:<br /> default_transaction_isolation = 'serializable'<br /></blockquote><p><br />Did you reload the configuration after making that change? SELECT<br />pg_reload_conf();<br /><br />Greetings,<br /><br />Andres Freund<br /></p></blockquote>

On Fri, Mar 8, 2019 at 11:11 AM PG Bug reporting form
<noreply(at)postgresql(dot)org> wrote:
>
> The following bug has been logged on the website:
>
> Bug reference: 15678

> P.S. If in the future you want to get more sensible error reports, then you
> should think about a few examples. People understand examples better than
> lengthy documentation descriptions.

Likewise a self-contained executable example makes diagnosing bug
reports considerably easier. As Andres said this is likely a caching
issue related to when changed settings take effect after making a
physical change.

David J.

<div xmlns="http://www.w3.org/1999/xhtml"><div style="color:rgb(0,0,0);font-family:Arial,sans-serif;font-size:15px;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;text-align:start;text-transform:none;white-space:normal;background-color:rgb(255,255,255);"><div>Problem solved.</div><div>The reason was that I used pgcli.</div><div>When I use psql, everything works as it should.</div><div>I assume that pgcli chooses transaction isolation level by itself.</div></div><div style="color:rgb(0,0,0);font-family:Arial,sans-serif;font-size:15px;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;text-align:start;text-transform:none;white-space:normal;background-color:rgb(255,255,255);"> </div><div style="color:rgb(0,0,0);font-family:Arial,sans-serif;font-size:15px;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;text-align:start;text-transform:none;white-space:normal;background-color:rgb(255,255,255);"><a target="_blank" style="color:rgb(153,0,153);" rel="noopener noreferrer" href="https://www.sql.ru/forum/1310005/funkciya-v-pg11-2-ne-zapuskaetsya-s-urovnem-izolyacii-serializable">https://www.sql.ru/forum/1310005/funkciya-v-pg11-2-ne-zapuskaetsya-s-urovnem-izolyacii-serializable</a></div><div style="color:rgb(0,0,0);font-family:Arial,sans-serif;font-size:15px;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;text-align:start;text-transform:none;white-space:normal;background-color:rgb(255,255,255);"> </div><div style="color:rgb(0,0,0);font-family:Arial,sans-serif;font-size:15px;font-style:normal;font-variant-ligatures:normal;font-variant-caps:normal;font-weight:400;text-align:start;text-transform:none;white-space:normal;background-color:rgb(255,255,255);">Thank you!</div></div><div><br /></div><div><br /></div><div>08.03.2019, 22:36, "David G. Johnston" &lt;david(dot)g(dot)johnston(at)gmail(dot)com&gt;:</div><blockquote type="cite"><p>On Fri, Mar 8, 2019 at 11:11 AM PG Bug reporting form<br />&lt;<a href="mailto:noreply(at)postgresql(dot)org">noreply(at)postgresql(dot)org</a>&gt; wrote:<br /></p><blockquote><br /> The following bug has been logged on the website:<br /><br /> Bug reference: 15678<br /></blockquote><p><br /></p><blockquote> P.S. If in the future you want to get more sensible error reports, then you<br /> should think about a few examples. People understand examples better than<br /> lengthy documentation descriptions.<br /></blockquote><p><br />Likewise a self-contained executable example makes diagnosing bug<br />reports considerably easier. As Andres said this is likely a caching<br />issue related to when changed settings take effect after making a<br />physical change.<br /><br />David J.<br /></p></blockquote>