I am new to zabbix and I am currently evaluating zabbix for monitoring data provided in snmp traps from a number of snmp agents. I have been able to receive the traps in zabbix and can see the default data for each host in the latest data page. I need to provide a couple of pieces of information from the trap, more than one OID value, for the users to see in the problem. I created a master item which correctly selects the traps of interest with regex. Next I create a few dependent items which extract the values from a couple of different OIDs, again using regex in preprocess of the item. So, for example, I have <trap>.details, <trap>.device, <trap>.device_ip, <trap>.severity from the dependent items.
In my trigger my expression compares the severity with an interger to determine the alarm condition, then check the device and device_ip bytelength for a value greater that zero in order to get those items into the trigger. I then set the trigger name to {ITEM.VALUE1}, and the operational data to {ITEM.VALUE2}:{ITEM.VALUE3} in the trigger.
However, when I see a new problem in the dashboard, the problem name comes from a different trap instance that do the operation data. It appears that the dependent items are independently processing different traps at different times, and the trigger expression includes values from different traps.
An example of one of my triggers:
Name:" Loss Of Input Sync {ITEM.VALUE3}"
Operation Data: "{ITEM.VALUE2}"
Expression:
last(/NMX Traps Template/loss-of-input-sync.severity)=1
and bytelength(last(/NMX Traps Template/loss-of-input-sync.device))>0
and bytelength(last(/NMX Traps Template/loss-of-input-sync.details))>0
In the problems from this trigger the Name and Operation Data contain data from two different traps.
Is there a setting that will force all the dependent items to process the same trap.??
Thanks for any help.
In my trigger my expression compares the severity with an interger to determine the alarm condition, then check the device and device_ip bytelength for a value greater that zero in order to get those items into the trigger. I then set the trigger name to {ITEM.VALUE1}, and the operational data to {ITEM.VALUE2}:{ITEM.VALUE3} in the trigger.
However, when I see a new problem in the dashboard, the problem name comes from a different trap instance that do the operation data. It appears that the dependent items are independently processing different traps at different times, and the trigger expression includes values from different traps.
An example of one of my triggers:
Name:" Loss Of Input Sync {ITEM.VALUE3}"
Operation Data: "{ITEM.VALUE2}"
Expression:
last(/NMX Traps Template/loss-of-input-sync.severity)=1
and bytelength(last(/NMX Traps Template/loss-of-input-sync.device))>0
and bytelength(last(/NMX Traps Template/loss-of-input-sync.details))>0
In the problems from this trigger the Name and Operation Data contain data from two different traps.
Is there a setting that will force all the dependent items to process the same trap.??
Thanks for any help.
Keep experimenting with Zabbix maybe it is hard at start but it is very flexible (IMO).
Comment