Hello,
I am trying to replace the default fs monitoring triggers by predictive ones like this:
"According to the data collected during the last 15 days, if the drive should reach 95% in less than a week, then trigger"
I translated it like this:
{DSIN - Template OS Linux:vfs.fs.size[{#FSNAME},pfree].timeleft(15d,,5)}>1w and {DSIN - Template OS Linux:vfs.fs.size[{#FSNAME},pfree].timeleft(15d,,5)}<>-1
It works, but if I install a new server, zabbix will base its prediction not on the last 15 days but on the last few measures it has, and of course, the prediction is wrong.
So I would like to had a condition to my trigger that would check if zabbix has enough data to get a decent prediction, like this:
"According to the data collected during the last 15 days, if the data from 15 days ago is available and the drive should reach 95% in less than a week, then trigger"
Is there a way to do this?
Thanks,
Arnaud
I am trying to replace the default fs monitoring triggers by predictive ones like this:
"According to the data collected during the last 15 days, if the drive should reach 95% in less than a week, then trigger"
I translated it like this:
{DSIN - Template OS Linux:vfs.fs.size[{#FSNAME},pfree].timeleft(15d,,5)}>1w and {DSIN - Template OS Linux:vfs.fs.size[{#FSNAME},pfree].timeleft(15d,,5)}<>-1
It works, but if I install a new server, zabbix will base its prediction not on the last 15 days but on the last few measures it has, and of course, the prediction is wrong.
So I would like to had a condition to my trigger that would check if zabbix has enough data to get a decent prediction, like this:
"According to the data collected during the last 15 days, if the data from 15 days ago is available and the drive should reach 95% in less than a week, then trigger"
Is there a way to do this?
Thanks,
Arnaud