This repo contains Zabbix template and zabbix-agent configuration for simple Vertica monitoring.
TODO: Cluster nodes discovery, separate performance monitoring for every node.
On your Vertica server:
sudo apt-get install python-pip
pip install --pre pytz
pip install vertica-python
Create Vertica user zabbix
with password SECRET
, create role monitoring
, then grant monitoring
to zabbix
and set it default for user.
Note that password you set here also must be written into userparameter_vertica.conf
.
CREATE ROLE monitoring;
CREATE USER zabbix IDENTIFIED BY 'SECRET';
GRANT monitoring TO zabbix;
ALTER USER zabbix DEFAULT ROLE monitoring;
Then connect to your database:
\c mydatabase
Grant neccessary privileges to monitoring
:
CREATE VIEW active_events_mon AS SELECT * FROM active_events;
CREATE VIEW cpu_usage_mon AS SELECT * FROM cpu_usage_mon;
CREATE VIEW database_backups_mon AS SELECT * FROM v_monitor.database_backups;
GRANT SELECT ON activee_events_mon TO monitoring;
GRANT SELECT ON cpu_usage_mon TO monitoring;
GRANT SELECT ON database_backups_mon TO monitoring;
Put vertica_stats.py
to directory where zabbix-agent can execute it, e.g. /home/zabbix/scripts
, ang chmod+chown it:
chown zabbix:zabbix vertica_stats.py
chmod 755 vertica_stats.py`.
Then put userparameter_vertica.conf
to your zabbix_agentd.d
directory, e.g. /etc/zabbix/zabbix_agentd.d/
and restart your zabbix-agent.