Maintenance window and suppressing restarted alerts

[mike smith]

we have a nightly maintenance window where the servers are shutdown overnight. at 8am, they start back, and my maintenance window closes at 8:30. my maintenance window is configured for NO data collection.

i get a problem/ok alert for every server stating it has just been restarted shortly after the window is closes.

i'm trying to suppress these from happening every morning

does anyone have an idea on how to accomplish that? i'd rather not disable that trigger (easy fix), but maybe add an additional expression to check about the time of day or time expired since last maintenance window?

[mike smith]

what i'm working on as a possible solution is adding two additional checks to the trigger as ANDs.

{Template OS Linux 8-6:system.uptime.change(0)}<0 & {Template OS Linux 8-6:system.localtime.time(0)}>121500 & {Template OS Linux 8-6:system.localtime.time(0)}<220000

[mike smith]

this is not working as expected.

https://www.zabbix.com/forum/showthread.php?t=10825

even that says i'm on the right path, but i'm still getting alerts on these machines for rebooting during a maintenance window.

anyone got any ideas?

[bbrendon]

What do you have for actions?

[mike smith]

Maintenance status not in maintenance
Trigger value = PROBLEM

[bbrendon]

I seem to remember having this problem. I'm not sure if it was ever resolved. We ended up changing our workflows to get around it.

https://www.zabbix.com/forum/showthread.php?t=39449

[mike smith]

same problem, damn.

after some testing today, this is working, with a minor issue. sometimes, the 2nd and 3rd checks are coming back with *UNKNOWN*, which then fails to test properly in the trigger.

the trigger is defined as: {Template OS Linux 8-6:system.uptime.change(0)}<0 & {Template OS Linux 8-6:system.uptime.time(0)}>124500 & {Template OS Linux 8-6:system.uptime.time(0)}<220000

my maintenance window ends at 123500 and is set to no data collection. the machines actually power on at 120000. by adding a little extra time in the trigger (124500) i hope that the *UNKNOWN* data values are fixed.