Audit management

2018-08-10 Thread dangal


Dear, I would like to ask you to see what you recommend
I manage a production database with a nominated user with the same
permissions as the postgres user SUPERUSER INHERIT CREATEDB CREATEROLE
REPLICATION
The client is asking us for an audit about this nominated user, could you
recommend any extension or form that you believe is the best solution,
I am currently testing pgaudit



--
Sent from: http://www.postgresql-archive.org/PostgreSQL-general-f1843780.html



Re: Audit management

2018-08-10 Thread dangal
Achilleas thanks for your answer
A query, the only possible way out is the postgres log?
I would like to be able to throw the audit output to a different place so
that I can not access the file



--
Sent from: http://www.postgresql-archive.org/PostgreSQL-general-f1843780.html



temporary files

2019-10-07 Thread dangal
Hello everyone, I have a question to see if you can help me, I have set
work_mem in 100 MB but I have the following in the pgbadger

Queries generating the most temporary files (N)

Count Total size   Min size  Max sizeAvg size
   58   3.24 GiB 57.15 MiB  57.15 MiB 57.15 MiB

How could it be that if you require 57 MB, 100 MB will not reach ?
thanks a lot



--
Sent from: https://www.postgresql-archive.org/PostgreSQL-general-f1843780.html




Re: temporary files

2019-10-08 Thread dangal
thank you very much you take for your time
We raised the work_mem to 130 mb and there was no more problem!
Now we are seeing to improve the query, it is complicated because it is
generated by a product that we have installed in production!



--
Sent from: https://www.postgresql-archive.org/PostgreSQL-general-f1843780.html




Hot_standby_feedback = on

2019-02-21 Thread dangal


Dear, I have a query to make you
As much as you enable hot_standby_feedback = on, queries continue to be
canceled, in addition to enabling this parameter you have to modify any of
these?
#max_standby_archive_delay = 30s
#max_standby_streaming_delay = 30s
#wal_receiver_status_interval = 10s
The idea that we have is to have a primary, a standby for high availability
and a second standby database in cascade from the replica to be able to
execute reports and not interfere with the other two databases



--
Sent from: http://www.postgresql-archive.org/PostgreSQL-general-f1843780.html