Trigger dependencies question

[fdavidcn]

Hi guys!

I have a little question with trigger dependencies. Following this simple (and classic) scenario:

Zabbix------R1------AP===============ST----R2----'Some devices"

1. Zabbix is directly connected to R1 via ethernet
2. R1 is directly connected to AP via ethernet
3. AP is connected to ST via radio
4. ST is directly connected to R2 via ethernet
5. R2 is directly connected to some Access Point (Sectors)

I have dependencies configured and working well when a device comes DOWN: (If AP goes offline i only receive one mail "AP device is OFFLINE"). At this point all is good.

The problem/question is when AP recovers. I receive one email for every "child" device (ST, R2, and Access Points...)

This situation is normal or i can do anything to avoid this?

I want this (if possible):
1. If AP goes DOWN, i want to receive only one mail (This is correct with actual configuration)
2. When AP goes UP, i want to receive only one mail: "Device AP is ONLINE again"

[B]Monitor configuration:[/B]
Name: Latency
Type: Simple Check
Monitor: icmppingsec
Interval: 5s

[B]Trigger configuration:[/B]
Name: Device is OFFLINE
Expression: {services:icmppingsec.avg(#4)}=0

OK Event generation: Expression
Problem Event: Single
OK Closes: All events
No TAGs are used

Zabbix Version: 3.4

Thanks!!

[kernbug]

Hi,

For me it was a little bit confusing, but you need to set 'Notify all involved': https://www.zabbix.com/documentation...ery_operations

[fdavidcn]

Hi,

For me it was a little bit confusing, but you need to set 'Notify all involved': https://www.zabbix.com/documentation...ery_operations

Hi kernbug, i will check this in documentation

Thanks!

[fdavidcn]

Hi kernbug,

I was reading this documentation and my server is configured as showed in the wiki (notify all involved) and this is ok but as i explained in previous post, i dont want to receive one email for every host behind affected node. Following initial example:

Zabbix------R1------AP===============ST----R2----AP1 & AP2

When AP comes OFFLINE and then ONLINE i receive theses emails:

1. Device AP is OFFLINE (This notification is Ok)
2. Device AP is ONLINE (This notification is Ok)

3. Device ST is ONLINE (i dont want this email)
4. Device R2 is ONLINE (i dont want this email)
5. Device AP1 is ONLINE (i dont want this email)
6. Device AP2 is ONLINE (i dont want this email)

It is posible to "re-check" child nodes after parent node goes ONLINE before the server activate their triggers?

For example, when AP goes ONLINE and child is still "OFFLINE", say to zabbix "Hey, wait a little and make a new probe before say that node is OFFLINE and if in the second probe still OFFLINE, then send notification".

Thanks!!