Getting problem and recovery notification at the same time

[bashman]

I get the problem and recovery notification at the same time, when I should only get the problem notification.

This is my action scenario:

Conditions:

(A) Host group = "Linux Servers"
(B) Host group = "Windows Servers"
(C) Trigger description not like "Inactive Zabbix Agent"
(D) Trigger severity >= "Average"
(E) Trigger value = "PROBLEM"
(F) Maintenance status not in "maintenance"

Escalation period:

86400

Operations:

1 - 0 Send message to Group "Linux Admins" Default Immediately

[Davidus]

I have same problem; in addition my triggers fire up even if they shouldn’t

[bashman]

This only ocurrs on this action, the rest are ok.

[Davidus]

Whats you trigger expression?

[bashman]

It happens with various triggers, all of them are similiar.

High	Enabled	Low free disk
Depends on :
linux5 : No Ping to linux5	({linux5:free.disk[/mnt/general].last(0)}=4G | {linux55:free.disk[/mnt/general].last(0)}<4G)	
 
Average	Enabled	Low free disk
Depends on :
linux5 : No Ping to linux5	({linux5:free.disk[/mnt/general].last(0)}=5G | {linux5:free.disk[/mnt/general].last(0)}<5G) & {linux5:[/mnt/general].last(0)}>4G

The items are external checks.

[Davidus]

http://www.zabbix.com/forum/showthread.php?t=22237

…might be helpful

[bashman]

Thanks for your post Davidus, but in my case I'm using the same item in my trigger expression.

The high severity trigger fires up and sends out the problem notification when "free.disk" item value is 3.19, but the average severety trigger sends out the ok notification.

The average severity trigger shouldn't send the ok notification because the item value is 3.19.

This problem only happens sometimes.

It seems to have some relation to https://support.zabbix.com/browse/ZBX-3777

[Davidus]

linux5 : No Ping to linux5	({linux5:free.disk[/mnt/general].last(0)}=5G | {linux5:free.disk[/mnt/general].last(0)}<5G) & {linux5:[/mnt/general].last(0)}>4G

Ok, I don’t understand this part of the trigger
& {linux5:[/mnt/general].last(0)}>4G

Shouldn’t before [/mnt/general] be some item name?

[bashman]

linux5 : No Ping to linux5	({linux5:free.disk[/mnt/general].last(0)}=5G | {linux5:free.disk[/mnt/general].last(0)}<5G) & {linux5:[/mnt/general].last(0)}>4G

Ok, I don’t understand this part of the trigger
& {linux5:[/mnt/general].last(0)}>4G

Shouldn’t before [/mnt/general] be some item name?

Yeah, that's right, it's:

linux5 : No Ping to linux5	({linux5:free.disk[/mnt/general].last(0)}=5G | {linux5:free.disk[/mnt/general].last(0)}<5G) & {linux5:free.disk[/mnt/general].last(0)}>4G

[Davidus]

Are you sure that the same Action sends both problem and recovery messages?
I had scenario where two different Action scenarios where sending messages if something was happening.

[bashman]

Yeah, I'm sure. It only happends sometimes.