Is there a way to make triggers more intelligent so they ignore scheduled activities on the server they are monitoring? For example, SQL backups are generating a lot of noise for disk read/write triggers. Can the triggers that usually go off when these backups are running be made to recognize that the activity being incurred at the time is consistent with previous time periods and thus, it should not alert on it? As I mentioned, that's one example. It is not necessarily just SQL backups. It could be some batch process that happens at the same time every day or week that causes a spike in activity, etc.
We looked at maintenance windows, but that approach doesn't work for us because it is an 'all or nothing' approach to silencing alerts and also won't scale well with our fleet size.
Same thing with defining 'less/more than time' trigger expressions for the same scale reasons. We'd like a more intelligent approach that doesn't involve hard coding time periods.
How have you solved for these?
We are using Zabbix 5.0 LTS​
We looked at maintenance windows, but that approach doesn't work for us because it is an 'all or nothing' approach to silencing alerts and also won't scale well with our fleet size.
Same thing with defining 'less/more than time' trigger expressions for the same scale reasons. We'd like a more intelligent approach that doesn't involve hard coding time periods.
How have you solved for these?
We are using Zabbix 5.0 LTS​