Zabbix threshold configuration

They are variable names to store different information

*trigger thresholds

*different filters

*credentials

*etc...

They have multiple levels, each new one overriding the previous

GLOBAL MACRO

{$PROC_NUM} = 200

TEMPLATE MACRO

{$PROC_NUM} = 300

HOST MACRO

{$PROC_NUM} = 500

Different servers may have different threshold values for items like CPU load, free memory size, number of processes etc

{Template OS: system.cpu.load[percpu,avg1].avg(5m)}>5

{Template OS: vm.memory.size[available].last(0)}<20M

{Template OS: proc.num[].avg(5m)}>300

Dynamic thresholds in templates

Use {$USER_MACROS} as threshold values for items like CPU load, free memory size, number of processes etc.

{Template OS: system.cpu.load[percpu,avg1].avg(5m)}>{$CPU_LOAD}

{Template OS: vm.memory.size[available].last(0)}<{$MEMORY_FREE}

{Template OS: proc.num[].avg(5m)}>{$PROC_NUM}

Different servers have different port numbers for tcp/udp ports like ssh, http, https etc.

{Template OS: net.tcp.service[ssh,22]

{Template OS: net.tcp.service[http,80]

{Template OS: net.tcp.service[https,443] 

Fixed thresholds for LLD (Low-Level Discovery) items

Different mount points will have different size, hence different low space warning thresholds

*/boot small size 100M

*/ medium size 10G

*/data large size 1TB