Hi,
I've setup Zabbix to monitor status and performance on network device. Its include around 20 routers and 60 switchs with around 20 items on each.
Some switch are stack together and I have some problem with dependency.
Like Switch B is connect to Switch A. Then Switch A is connect to BackBone.
I set trigger on switch A :
- Trigger 1 - Disaster - Switch A is unreachable.
- Trigger 2 - Warning - Uplink interface is DOWN.
- Trigger 3 - Warning - Downlink interface is DOWN.
- Trigger 4 - Information - Uplink traffic on router A has reach 90% on last 5 minutes.
- Trigger 5 - Information - Uplink error on router A has reach more than 2 fps on last 60 seconds.
- Trigger 6 - Information - CPU overload (more than 95%) on last 5 minutes.
I set trigger on switch B :
- Trigger 7 - Disaster - Switch B is unreachable.
- Trigger 8 - Warning - Uplink interface is DOWN.
- Trigger 9 - Information - Uplink traffic on router A has reach 90% on last 5 minutes.
- Trigger 10 - Information - Uplink error on router A has reach more than 2 fps on last 60 seconds.
- Trigger 11 - Information - CPU overload (more than 95%) on last 5 minutes.
With these trigger, I can advise by e-mail when the switch is down or unreachable. I also advise about the status on interfaces (Uplink, Downlink). I use "Information" status to evaluate the global load of the equipement in order to upgrade link/equipement when its needed.
Here the dependency I want.
Trigger 11 (CPU overload on switch B) depend on Trigger 7.
Trigger 10 (Error threshold on switch B) depend on Trigger 7.
And so on like every performance trigger on a switch depend of the availability on the switch. I do it on both switch A and B.
Then,
Trigger 7 (Switch B status) depend on the Trigger 1 (Swith A status).
With that, I try to reduce the amount of false TRUE when a equipement get down and other equipement behind been unreachable but its still working. For example, I got problem on the switch where the Zabbix server is connected. Then, instead of getting just the faulty switch trigger goes ON, you will get all you trigger goes ON.
But when I add the dependency between Trigger 7 and Trigger 1, I receive an SQL error :
SQL error: Duplicate entry '12280-12225' for key 1
Query: insert into trigger_depends (triggerid_down,triggerid_up) values (12280,12225)
The dependency looks like to be add, other dependency also looks like to be update, but I'm not sure if its work good.
Thanks for geting me information about this error message.
ITou.
I've setup Zabbix to monitor status and performance on network device. Its include around 20 routers and 60 switchs with around 20 items on each.
Some switch are stack together and I have some problem with dependency.
Like Switch B is connect to Switch A. Then Switch A is connect to BackBone.
I set trigger on switch A :
- Trigger 1 - Disaster - Switch A is unreachable.
- Trigger 2 - Warning - Uplink interface is DOWN.
- Trigger 3 - Warning - Downlink interface is DOWN.
- Trigger 4 - Information - Uplink traffic on router A has reach 90% on last 5 minutes.
- Trigger 5 - Information - Uplink error on router A has reach more than 2 fps on last 60 seconds.
- Trigger 6 - Information - CPU overload (more than 95%) on last 5 minutes.
I set trigger on switch B :
- Trigger 7 - Disaster - Switch B is unreachable.
- Trigger 8 - Warning - Uplink interface is DOWN.
- Trigger 9 - Information - Uplink traffic on router A has reach 90% on last 5 minutes.
- Trigger 10 - Information - Uplink error on router A has reach more than 2 fps on last 60 seconds.
- Trigger 11 - Information - CPU overload (more than 95%) on last 5 minutes.
With these trigger, I can advise by e-mail when the switch is down or unreachable. I also advise about the status on interfaces (Uplink, Downlink). I use "Information" status to evaluate the global load of the equipement in order to upgrade link/equipement when its needed.
Here the dependency I want.
Trigger 11 (CPU overload on switch B) depend on Trigger 7.
Trigger 10 (Error threshold on switch B) depend on Trigger 7.
And so on like every performance trigger on a switch depend of the availability on the switch. I do it on both switch A and B.
Then,
Trigger 7 (Switch B status) depend on the Trigger 1 (Swith A status).
With that, I try to reduce the amount of false TRUE when a equipement get down and other equipement behind been unreachable but its still working. For example, I got problem on the switch where the Zabbix server is connected. Then, instead of getting just the faulty switch trigger goes ON, you will get all you trigger goes ON.
But when I add the dependency between Trigger 7 and Trigger 1, I receive an SQL error :
SQL error: Duplicate entry '12280-12225' for key 1
Query: insert into trigger_depends (triggerid_down,triggerid_up) values (12280,12225)
The dependency looks like to be add, other dependency also looks like to be update, but I'm not sure if its work good.
Thanks for geting me information about this error message.
ITou.