Zabbix server is not running: the information displayed may not be... (split servers)

[Jaceqp]

Hi there.
I'm trying to troubleshoot my split zabbix environment. I've had everything up'n'running on all-in-one zabbix srv. Then I created separate servers for frontend and DB, created empty DB and restored backup into it. I guess I tweaked all conf files as expected, reconfigured GUI using IP/zabbix/config.php - went smoothly, no issues connecting to DB...

Anything else to check?

zabbix 7.4.2 + postgresql 17


PS. Here's some issue connecting to DB in logs (kinda weired since gui configurator shows no issues):

801:20260410:100002.350 database is down: reconnecting in 10 seconds
801:20260410:100012.366 [Z3001] connection to database 'zabbix' failed: [0] FATAL:  password authentication failed for user "zabbix"

[Jaceqp]

Allright first update... It seems I had a typo in zabbix_server.conf in db password.
Now /var/log/zabbix/zabbix_server.log shows something new:

5545:20260410:102910.144 cannot use database "zabbix": database is not a Zabbix database
5620:20260410:102922.357 Starting Zabbix Server. Zabbix 7.4.2 (revision 7aa4e0782fe).
5620:20260410:102922.357 ****** Enabled features ******
5620:20260410:102922.357 SNMP monitoring: YES
5620:20260410:102922.358 IPMI monitoring: YES
5620:20260410:102922.358 Web monitoring: YES
5620:20260410:102922.358 VMware monitoring: YES
5620:20260410:102922.358 SMTP authentication: YES
5620:20260410:102922.358 ODBC: YES
5620:20260410:102922.358 SSH support: YES
5620:20260410:102922.358 IPv6 support: YES
5620:20260410:102922.358 TLS support: YES
5620:20260410:102922.358 ******************************
5620:20260410:102922.358 using configuration file: /etc/zabbix/zabbix_server.conf
5620:20260410:102922.384 [Z3005] query failed: [0] PGRES_FATAL_ERROR:ERROR: relation "users" does not exist
LINE 1: select userid from users limit 1
^
[select userid from users limit 1]

[cyber]

verify that your db restore succeeded... seems like you have empty DB there...

[Jaceqp]

What I did was:

- create local db dump (downer server):

# pg_dump -h [ZABBIX SERVER] -U [ZABBIX DB USER] > /tmp/zabbix_dump.sql

- moved dump file to destination server
- removed all temporary/test zabbix postgres DB's using 'drop database xxx'
- created new empty DB named zabbix:

postgres=# create database zabbix;

- restored dump file on a DB:

# sudo -u postgres psql -dzabbix -f /tmp/zabbix_dump.sql

- restarted all related services like zabbix-server, agent, postgresql etc.
- using http://zabbix-server-ip/zabbix/setup.php reconfigured initial setup (went flawlessly connecting to DB zabbix).
Web gui loads, login screen accepts credentials (assuming stored in DB anyway), I can see created custom dashboards, maps etc yet with warning 'zabix server is not running' at the bottom...
What am I missing?

[cyber]

If you split up a all-in-one server to 3 separate ones (db, frontend, server), then you are probably missing a connection from frontend server to zabbix server 10051 port.. Frontend does verify, if backend is running..

[Jaceqp]

Frontend + server is one server. Only DB is being moved to another one. I think I got it solved cause I had both db name and db schema variables set in zabbix-server conf file. I hashed out dbschema and that solved the issue.
Now I'm working on "PostgreSQL: service is down" on zabbix server_DB host itself. Seems triggers/template issue...