I'm new on the forum, so greetings to everyone.
ZABBIX is very good monitoring system and we are very satisfied with it. We migrated few weeks ago from Nagios and it was a good decision :-)
I'm pretty sure that you know S.M.A.R.T. feature of modern hard drives. S.M.A.R.T. is a standard for monitoring various attributes of hard disk drives, such as read failures, temperatures, speeds, etc. Using these attribute values, you can predict hard drive failures in approximately 30% cases.
Not all monitored values are important and moreover several of them are vendor specific. However there are some of them, which are certainly important, implemented by most manufacturers, and can help to detect broken hard disk before its fatal failure.
Here is a command which extracts these important values from "smartctl" output (device is /dev/sda):
We would like to have this in ZABBIX, for every our machine, for every installed drive. Also other parameters, such as temperature could be handy, as you can watch that cute graphs and see the time point where temperature started to go up or down, and so on.
My question is, if something like this is already implemented. If so, please point me to the resource. I tried to find, but without any luck.
If this is not implemented, we will be happy to do this as an our contribution to the Zabbix development. Hopefully it will be welcomed. However, I have no idea how to develop "plugin" for Zabbix, I do not know where to start. Please point me to some developer resource. Comments on implementation, what we should aware before we start to do this, are very appreciated.
Thank you.
ZABBIX is very good monitoring system and we are very satisfied with it. We migrated few weeks ago from Nagios and it was a good decision :-)
I'm pretty sure that you know S.M.A.R.T. feature of modern hard drives. S.M.A.R.T. is a standard for monitoring various attributes of hard disk drives, such as read failures, temperatures, speeds, etc. Using these attribute values, you can predict hard drive failures in approximately 30% cases.
Not all monitored values are important and moreover several of them are vendor specific. However there are some of them, which are certainly important, implemented by most manufacturers, and can help to detect broken hard disk before its fatal failure.
Here is a command which extracts these important values from "smartctl" output (device is /dev/sda):
Code:
smartctl -A /dev/sda | grep -E -i '^[ ]*(1|5|196|197|198|201|220)[ ]'
My question is, if something like this is already implemented. If so, please point me to the resource. I tried to find, but without any luck.
If this is not implemented, we will be happy to do this as an our contribution to the Zabbix development. Hopefully it will be welcomed. However, I have no idea how to develop "plugin" for Zabbix, I do not know where to start. Please point me to some developer resource. Comments on implementation, what we should aware before we start to do this, are very appreciated.
Thank you.
Comment