Hello!

Sometimes we have to use "Current User ID", and "User Name" in the
Triggers to make a log into a table. These values are based on our User,
not in the PSQL role.

Now we use a temporary table to do this.
When the user logged into the application, we created a temporary table
with the same name (user_info) and structure. This holds the data (id,
name, machine info, ip address).

In the trigger we try to find this table (in the LOCAL_TEMPORARY schema).
Then we read the row into a JSON record, and then into PLPGSQL variables.
Tables can exist with the same name, so this is the safest solution.
If the User ID is invalid (none or empty) that means this is a background
operation, and then we don't need to log the changes.

But maybe there is a better way to somehow store some session based data
and use it in the triggers.
Because if these selects are slow, the trigger is also slow. So when I
start an UPDATE command in a big table, maybe this slows down the whole
operation.

Note:
A table with the PID key is not enough, because the PID is a repeated
value.
I logged it and in the Windows system there are many of the same values
(10001, 10004, etc.).
Ok, I can combine with session creation time. But for this I also need to
start a select in the pgstatactivty table.

So maybe you have an easier way to point to a record in a session.
Important: the PG servers are different, the lesser version is 11, and we
have only a Database Owner role. We can't configure the server.

What is your opinion? Is there any way to get session based data?
As I read before, we can't set the session variables onfly.

SET
┌─────────────────┐
│ current_setting │
╞═════════════════╡
│ hello           │
└─────────────────┘
(1 row)

Best regards
dd

Hello!

Sometimes we have to use "Current User ID", and "User Name" in the > Triggers to make a log into a table. These values are based on our User, > not in the PSQL role.

Now we use a temporary table to do this.

When the user logged into the application, we created a temporary table > with the same name (user_info) and structure. This holds the data (id, > name, machine info, ip address).

In the trigger we try to find this table (in the LOCAL_TEMPORARY schema).

Then we read the row into a JSON record, and then into PLPGSQL variables.
Tables can exist with the same name, so this is the safest solution.

If the User  ID is invalid (none or empty) that means this is a > background operation, and then we don't need to log the changes.

But maybe there is a better way to somehow store some session based data > and use it in the triggers.
Because if these selects are slow, the trigger is also slow. So when I > start an UPDATE command in a big table, maybe this slows down the whole > operation.

Note:
A table with the PID key is not enough, because the PID is a repeated > value.
I logged it and in the Windows system there are many of the same values > (10001, 10004, etc.).
Ok, I can combine with session creation time. But for this I also need > to start a select in the pgstatactivty table.
So maybe you have an easier way to point to a record in a session.
Important: the PG servers are different, the lesser version is 11, and > we have only a Database Owner role. We can't configure the server.
What is your opinion? Is there any way to get session based data?

As I read before, we can't set the session variables onfly.

CREATE OR REPLACE FUNCTION public.session_test()
 RETURNS void
 LANGUAGE plpgsql
AS $function$
DECLARE
_test_var varchar := current_setting('test.session_var', 't');
BEGIN
RAISE NOTICE 'Variable is %', _test_var;
END;
$function$

--------------

BEGIN
test=*# set local test.session_var = 'test';
SET
test=*# select session_test();
NOTICE:  Variable is test
 session_test
--------------

Best regards

dd

-- Adrian Klaver