After maintenance period

[avecsi]

Hi All,

I was recently upgraded my zabbix to 2.2 from 2.0

If I put a server into maintenance everything work as should be.
No disaster, no email if I reboot the server, but if when the maintenance period ended or I remove the server from the Host and Groups

I got a lot OK email

I dont need this "OK" emails, I know that it is everything fine thats why I remove the server(s) from the Host and Groups part.

In my maintenance periods I always collecting data.

Is that a known bug or should I just select no data collection for maintenance type
?

Thanks,
Andrew

[ingus.vilnis]

Hello Andrew,

Please check that you have disabled your Recovery message in Configuration -> Actions

https://www.zabbix.com/documentation/2.2/manual/config/notifications/action

Best Regards,
Ingus

[avecsi]

Hello Ingus,

Yes the recovery message is turned off.

My Action:

https://drive.google.com/folderview?id=0ByLO-LIP5KHfbjlhdl9XdThIeGM&usp=drive_web

The default message is not empty. For sec I removed the content.

I used this settings in the 2.0.2 version.

My current version is 2.2.3

Thanks,
Andrew

[ingus.vilnis]

Andrew,

Thank you for the information!

Please add the following line to your Action conditions:

Trigger value = PROBLEM

Thus you will get notified only when trigger goes into PROBLEM state and not to OK state.

Best Regards,
Ingus

[avecsi]

I need notification both state

I cannot turn of the OK status

[ingus.vilnis]

Andrew,

I'm sorry, I don't get the idea then.

Also what do you mean by saying this

I remove the server(s) from the Host and Groups part.

What is the purpose of doing it?

https://www.zabbix.com/documentation/2.2/manual/maintenance

Best Regards,
Ingus

[avecsi]

We are using 3 types of Maintenance

1. Fix periods with fix servers
In this case we never change the periods or the servers
/like wsus automated upgrades at night at THU/

2. Fixed periods with notfixed server
In this case the periods is always the same but we always changing the server
/like Monday night maintenance - upgrade, reboot etc/

3. Emergency maintenance / deployments
Not fixed periods not fixed servers


My problem is that when a maintenance period ends or I remove a server from the maintenance I got a lot of only OK alerts. Even if the server is up and running nearly 20-50 minutes and there is no error.

We always collecting data for Monthly reports. So I cannot turn that off.

Are we using it in a wrong way?

[ingus.vilnis]

Hi Andrew,

So does this mean that in cases 2 and 3 you have periods where you simply add and remove hosts from the maintenance settings?

Not sure this is related but what do you have for "Type of calculation" Action Conditions tab?

Another thing that came to my mind is that you could add Acknowledgment status in Operation conditions. (see bottom of picture 3 in your previously added screenshots)

Best Regards,
Ingus

[avecsi]

Hi,

So does this mean that in cases 2 and 3 you have periods where you simply add and remove hosts from the maintenance settings?

Yes.

Not sure this is related but what do you have for "Type of calculation" Action Conditions tab?

Where can I find this?

Another thing that came to my mind is that you could add Acknowledgment status in Operation conditions. (see bottom of picture 3 in your previously added screenshots)

We are not using ACK option

[ingus.vilnis]

Not sure this is related but what do you have for "Type of calculation" Action Conditions tab?

Where can I find this?

Picture 2 here https://drive.google.com/folderview?...&usp=drive_web

So does this mean that in cases 2 and 3 you have periods where you simply add and remove hosts from the maintenance settings?

Yes.

I remember something similar some time ago. I cannot find the correct case but the conclusion there was that it is not really the best practice to throw the hosts simply in and out of one long maintenance period. It would be better to set up a specific maintenance window and then simply wait for it to end. Is that acceptable solution for you?

Best Regards,
Ingus

[avecsi]

Oh, yes I found.

(A) and (B or C) and (D)

Label Name
(A) Maintenance status not in maintenance
(B) Trigger value = PROBLEM
(C) Trigger value = OK
(D) Trigger severity = Disaster

I already added B C lines, but I think it is the same as
(A) and (B)
(A) Maintenance status not in maintenance
(B) Trigger severity = Disaster

That could be a workaround if i change the Periods