Hi all,
I am running Zabbix 2.2.2 using the precompiled binaries for RHEL6 and using postgres as the database.
I'm trying to monitor the status of a RHEL6 guest and display the status visually on a map. I've got 4 different applications defined (CPU, RAM, Disk, and Network Utilization) with several items assigned to each application. I've got triggers assigned to each item (i.e., CPU utilization exceeds 80%).
What I'd like is a map with 4 elements on it corresponding to the 4 applications for the RHEL6 guest. If a trigger associated with the CPU Utilization is triggered, the CPU element goes into a problem state. The problem is that it appears you can only link a single trigger to an element.
I've got a couple workarounds, but they are ugly. First would be to create a new map with all of the CPU triggers as elements and then set the original map element to type map and reference the new map. Do this four times, and voila. Except now I've got 5 times the number of maps per RHEL6 guest and this becomes extremely unmanageable the more guests you monitor.
Second workaround would be to split the host up into four hosts, all referencing the same hostname, one for each application and then tie each map element to each of the four hosts. This is a bit more manageable, but uglies up the dashboard and hosts screens with 4 times the number of hosts. I'm sure it also increases server and network load.
Is there anything I missed? A trigger group would be fantastic, or even just being able to tie multiple triggers to a map element. Anyone have any other workarounds?
Thanks in advance for any help.
rt
I am running Zabbix 2.2.2 using the precompiled binaries for RHEL6 and using postgres as the database.
I'm trying to monitor the status of a RHEL6 guest and display the status visually on a map. I've got 4 different applications defined (CPU, RAM, Disk, and Network Utilization) with several items assigned to each application. I've got triggers assigned to each item (i.e., CPU utilization exceeds 80%).
What I'd like is a map with 4 elements on it corresponding to the 4 applications for the RHEL6 guest. If a trigger associated with the CPU Utilization is triggered, the CPU element goes into a problem state. The problem is that it appears you can only link a single trigger to an element.
I've got a couple workarounds, but they are ugly. First would be to create a new map with all of the CPU triggers as elements and then set the original map element to type map and reference the new map. Do this four times, and voila. Except now I've got 5 times the number of maps per RHEL6 guest and this becomes extremely unmanageable the more guests you monitor.
Second workaround would be to split the host up into four hosts, all referencing the same hostname, one for each application and then tie each map element to each of the four hosts. This is a bit more manageable, but uglies up the dashboard and hosts screens with 4 times the number of hosts. I'm sure it also increases server and network load.
Is there anything I missed? A trigger group would be fantastic, or even just being able to tie multiple triggers to a map element. Anyone have any other workarounds?
Thanks in advance for any help.
rt
Comment