Hello,
I'm using the template Supermicro Aten by SNMP to monitor many Supermicro servers in Zabbix 7.4.3.
This template does an SNMP discovery of all temperature sensors with this OID:
discovery[{#SNMPVALUE},1.3.6.1.4.1.21317.1.3.1.2,{#SENSOR_DE SCR},1.3.6.1.4.1.21317.1.3.1.13]
This discovers CPU temperatures, NIC temperatures, and GPU temperatures. The problem is that all of these use the same {$TEMP_WARN} and {$TEMP_CRIT} macro; and the GPUs can run much hotter than the CPU or NIC - around 80*C in some cases. To avoid alarms I've had to set {$TEMP_CRIT} very high, higher than I would like for the CPUs.
Is there a way to have a different limit based on the item name? In the Discovery Rule Override I think I can change the severity based on {#SENSOR_DESCR}, but not the trigger value. Items get names like CPU 1 Temp, GPU1 Temp, etc.
Thanks,
-Jason
I'm using the template Supermicro Aten by SNMP to monitor many Supermicro servers in Zabbix 7.4.3.
This template does an SNMP discovery of all temperature sensors with this OID:
discovery[{#SNMPVALUE},1.3.6.1.4.1.21317.1.3.1.2,{#SENSOR_DE SCR},1.3.6.1.4.1.21317.1.3.1.13]
This discovers CPU temperatures, NIC temperatures, and GPU temperatures. The problem is that all of these use the same {$TEMP_WARN} and {$TEMP_CRIT} macro; and the GPUs can run much hotter than the CPU or NIC - around 80*C in some cases. To avoid alarms I've had to set {$TEMP_CRIT} very high, higher than I would like for the CPUs.
Is there a way to have a different limit based on the item name? In the Discovery Rule Override I think I can change the severity based on {#SENSOR_DESCR}, but not the trigger value. Items get names like CPU 1 Temp, GPU1 Temp, etc.
Thanks,
-Jason
Comment