Re: replication strange behavior - Mailing list pgsql-general

From Ron Johnson
Subject Re: replication strange behavior
Date
Msg-id CANzqJaB+FXug104Q=uCw+0cPAkA-VN7LtEtn=p0yz8xxUrryvg@mail.gmail.com
Whole thread Raw
In response to replication strange behavior  (Atul Kumar <akumar14871@gmail.com>)
Responses Re: replication strange behavior
List pgsql-general
On Thu, Nov 30, 2023 at 3:41 PM Atul Kumar <akumar14871@gmail.com> wrote:
Hi,

I have postgres 12 running in centos 7.

I have configured streaming replication between one master and one standby server.

In the pg_hba.conf file of the master server, I have put the standby server's hostname instead of IP and due to which replication got broken and I started getting below error.

FATAL:  no pg_hba.conf entry for replication connection from host "10.20.8.22", user "replication", SSL off

2023-11-30 12:00:25 PST|pid=24096|FATAL:  XX000: could not connect to the primary server: FATAL:  no pg_hba.conf entry for replication connection

from host "10.20.8.22", user "replication", SSL on


I put the ip of standby back in pg_hba.conf file of the master server and replication got started, though in the standby server I mentioned the hostname of master in primary_conninfo on the standby side.


So why is the master not reading the hostname of standby ? What am I missing here ?

 
You'd have to show us pg_hba.conf and /etc/hosts.

My guess, though, is that you're not taking the domain name into account.
 

pgsql-general by date:

Previous
From: Atul Kumar
Date:
Subject: replication strange behavior
Next
From: Adrian Klaver
Date:
Subject: Re: Two started cluster on the same data directory and port