Currently, we try to switch from Zabbix 6.2 to 7.0-2 for our new clusters (i.e. those do not have Zabbix 6.2 preinstalled, but we install 7.0-2 directly). However, while 6.2 runs without any visible issues, 7.0-2’s server stops after start. I am primarily looking for ways to debug this problem further, but if someone has straight up a solution, I am all there fore it.

The log indicates that it is a dns issue:

# /var/log/zabbix/zabbix_server.log
thread started [discovery worker #4]
cannot initialize asynchronous DNS library
cannot initialize asynchronous DNS library
One child process died (PID:1650808,exitcode/signal:1). Exiting ...
HA manager has been paused
zabbix_server [1650753]: Error waiting for process with PID 1650808: [10] No child processes
HA manager has been stopped
syncing trend data...
syncing trend data done
Zabbix Server stopped. Zabbix 7.0.0 (revision 49955f1fb5c).

We do use dnsmasq but had no issues with Zabbix 6.2. Name resolution via dnsmasq works in general.

In order to install Zabbix we make use of an Ansible playbook:

- name: Add zabbix repositories
  apt:
    deb: "https://repo.zabbix.com/zabbix/7.0/{{ ansible_distribution | lower }}
    /pool/main/z/zabbix-release/zabbix-release_7.0-2%2B{{ ansible_distribution | lower }}
    {{ ansible_distribution_version }}_all.deb"
    state: present
    force: true

# -------------------
# -- Zabbix Server --
# -------------------

- name: Debian based OS (Install database & Zabbix server)
  apt:
    name:
      - mariadb-server
      - zabbix-server-mysql
      - zabbix-sql-scripts
    state: present
    update_cache: true

- name: Install python modul 'PyMySQL'
  pip:
    name: PyMySQL

- name: Create zabbix database
  mysql_db:
    name: "{{ zabbix_conf.db }}"
    encoding: 'utf8'
    state: present
    login_user: root
    login_unix_socket: /run/mysqld/mysqld.sock

- name: Create zabbix database user
  mysql_user:
    name: "{{ zabbix_conf.db_user }}"
    password: "{{ zabbix_conf.db_password }}"
    priv: '{{ zabbix_conf.db }}.*:ALL'
    login_user: root
    login_unix_socket: /run/mysqld/mysqld.sock

- name: Check if zabbix schema exists
  failed_when: false
  shell: "echo describe users | mysql --user={{ zabbix_conf.db_user }} --password={{ zabbix_conf.db_password }} zabbix"
  changed_when: false
  register: zabbix_schema_exists
  tags:
    - skip_ansible_lint

- name: Import initial db schema
  shell: "set -o pipefail && zcat /usr/share/zabbix-sql-scripts/mysql/server.sql.gz | 
  mysql --user={{ zabbix_conf.db_user }} --password={{ zabbix_conf.db_password }} zabbix"
  when: zabbix_schema_exists.rc == 1
  args:
    executable: bash
  tags:
    - skip_ansible_lint

- name: Update Admin password
  mysql_query:
    login_db: "{{ zabbix_conf.db }}"
    login_user: "{{ zabbix_conf.db_user }}"
    login_password: "{{ zabbix_conf.db_password }}"
    query: update users set passwd='{{ zabbix_conf.admin_password | password_hash("bcrypt") }}' where username='Admin'

- name: Adjust zabbix server configuration
  template:
    src: zabbix/zabbix_server.conf.j2
    dest: /etc/zabbix/zabbix_server.conf
    mode: 0644
  notify: zabbix-server

- name: Start and Enable zabbix-server
  systemd:
    name: zabbix-server
    state: started
    enabled: true

Additional

There are more tasks running (like setting up the zabbix-agent and a web view), but those are unrelated to the problem, I think.

I have omitted the tags Ansible and dnsmasq on this question for now, because narrowing down on the problem is probably just a zabbix question and I don’t want to discourage potential answerers by adding a tag they might be unfamiliar with.