Resolved Backup error with Database

[papahausa]

Server operating system version

AlmaLinux 9.6 (Sage Margay)

Plesk version and microupdate number

Plesk Obsidian 18.0.71 Update #2

Hello.
When i make backup, i got error.

Warning: Database "papa.hausa_authelp"
Unable to make database dump. Error: Failed to exec mariadb-dump: Exit code: 2: WARNING: option --ssl-verify-server-cert is disabled, because of an insecure passwordless login.mariadb-dump: Got error: 2002: "Can't connect to server on 'localhost' (99)" when trying to connect

 

[Sebahat.hadzhi]

Hello, @papahausa . Could you please confirm the MariaDB version on your server and share the content of the /etc/my.cnf file?

 

[papahausa]

Hello, @papahausa . Could you please confirm the MariaDB version on your server and share the content of the /etc/my.cnf file?

[root@salmon ~]# cat /etc/my.cnf
#
# This group is read both both by the client and the server
# use it for options that affect everything
#
[client-server]

#
# include all files from the config directory
#
!includedir /etc/my.cnf.d

[mysqld]
sql_mode=ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION
bind-address = 127.0.0.1
local-infile=0

 

[Sebahat.hadzhi]

Thank you. Please try commenting out:

#bind-address = 127.0.0.1

then kindly let us know if the dump process still fails.

 

[Andreas Mause]

@papahausa And? Was this the solution? xxx.xxx.xxx.xxx -> public IP.

We have the same error. Do we need the bind-adress for replication? Strange, a manual backup worked yesterday.

Unable to make database dump. Error: Failed to exec mariadb-dump: Exit code: 2: WARNING: option --ssl-verify-server-cert is disabled, because of an insecure passwordless login.mariadb-dump: Got error: 2002: "Can't connect to server on 'localhost' (111)" when trying to connect
[client-server]
# Port or socket location where to connect
# port = 3306
socket = /run/mysqld/mysqld.sock

# Import all .cnf files from configuration directory
!includedir /etc/mysql/conf.d/
!includedir /etc/mysql/mariadb.conf.d/
[mysqld]
#Replication Master
server-id = 1
binlog_do_db = domain1_tld_de
binlog_do_db = domain2_tld_de
log_bin = log_bin = mysql-bin
expire_logs_days = 3
sync_binlog = 1
bind-address = xxx.xxx.xxx.xxx
binlog_format=ROW
local-infile=0

sql_mode=ERROR_FOR_DIVISION_BY_ZERO,NO_AUTO_CREATE_USER,NO_ENGINE_SUBSTITUTION

 

[Andreas Mause]

... okay, we NEED the puplic IP for replication. So, what can we do?

Kind regards,
Andreas

 

[Andreas Mause]

... it's only happens on the both databases domain1_tld_de and domain2_tld_de - the domains for replication.

 

[Andreas Mause]

... maybe we should note:

bind-address = 127.0.0.1, xxx.xxx.xxx.xxx

or

bind-adress = 0.0.0.0

It's a production server, so we can't test it properly. It is just surprising why it works for other domains on the server, but not for the two that are replicated.

 

[Andreas Mause]

Info: A manual backup produces the same problem.

 

[Andreas Mause]

Question: On the old server, "have_ssl" is setting to "No". Maybe a resolution?

 

[Sebahat.hadzhi]

Hi, @Andreas Mause . I don't believe the SSL setting has any relation to the reported error. The primary thing is that the connection cannot be established on the specified address. I am assuming that

bind-address = xxx.xxx.xxx.xxx

represents external IP address, so that's very likely the reason for the failure.

 

[Andreas Mause]

Thx for your feedback! But on the old Server (Ubuntu 22/MariaDB 10), bind-address was also set to the extern IP of this server (for replication reason). Everything was fine.

 

[Andreas Mause]

bind-address = 0.0.0.0 would be the solution - maybe. I will try it tonight ...

 

[Andreas Mause]

... problem resolved (bind-adress=0.0.0.0)! The firewall was adjusted accordingly ...