Action/Escalations Configuration

[Gsmith]

Hello,
I’m fairly new at creating Actions/Escalations.
I’m running CentOS7.3 with Zabbix Server 3.2.3 installed.
I’m trying to create an action for Zabbix agent to notify me that it is unavailable immediately, then every hour to all admin users when it Not Acknowledge.
After 1 Day there is an escalations that gets sent to CTO/Boss, this will continue till it’s acknowledged.

If the problem is acknowledge, BUT still is a problem I need a notification every day stating it’s still a problem i.e. not fixed.

What I tried was to create an action that has only two steps using Zabbix Agent Unavailable trigger.

Action Name = zabbix_agentd – Availability
Conditions using the Trigger=Zabbix-agent Unavailable

Operations Step duration= 3600s
1. Operation steps = 1-0 immediately duration=Default Send to all Admin Users
Operation Conditions = Event Acknowledged=Not Ack
2. Operation steps = 25-0 1Day, duration=86400 Send to CTO (i.e. Boss)
Operation Conditions = none
During my testing everything seems to work fine, it sends email immediately that there is a problem and every hour to all Admin users. Also after one day it send an email to my CTO/Boss. The problem that occurs is no matter when you acknowledge the problem (couple hours or after a day) my CTO still get an email every hour. It looks like this is coming from the default settings on the Operation part, but I thought I configured it to only send once a day. I’ve have tried different configuration such as adding the second step with;
Operation Conditions = Event Acknowledged=Not Ack
This added configuration stop send emails alerts to CTO after it was acknowledged, but no reminder emails that there is still a problem. I have been at this and all over the web, Zabbix forums and

https://www.zabbix.com/documentation/3.2/manual/config/notifications/action/escalations

Any Help or advice would be appreciated
Thank you

[Gsmith]

Testing New Config.

Old Action/Operation
Default step duration is 1 hour (3600 Sec)
Steps;
1. 1-0 Alert Immediately Users. Alert every Hour, Conditions=Not-Acknowledge
2. 25-0 After 24 hours Escalate alert to Sysadmin and users, 1 day, 00:00:00 86400, Condition= none

If alert is acknowledged a reminder alert will be sent every 24 hours if problem still exists.
PROBLEM: After 24 hours the Alert has been acknowledge it still goes off every hour (i.e. default operation step duration).

New Action/Operation
Default step duration is 1 hour (3600 Sec)
Steps;
1. 1-0 Alert Immediately Users. Alert every Hour, Conditions=Not-Acknowledge
2. 25-0 After 24 hours Escalate alert to Sysadmin and users, 1 day, 00:00:00 86400, Condition= Not-Acknowledge
3. 25 -0 After 24 hours Escalate alert to Sysadmin and Admin users, 1 day, 00:00:00 86400, Condition= Acknowledge
This worked in the testing stage, I was using 10 minutes= 1 hour and 30 minutes = 1 day.
After 1 hour of alerts I acknowledge the problem, which would equal 2 days of not-acknowledging problem.
I stopped getting alert every 10 minutes and only received alerts to Sysadmin every thirty minutes.

Second test I’m using the correct time duration with an added step and different configuration in step 2 as shown below;
Default step duration is 1 hour (3600 Sec)
Steps;
1-0 Alert Immediately Users. Alert every Hour, Conditions=Not-Acknowledge
25-0 After 24 hours Escalate alert to Sysadmin and users, 1 day, 00:00:00 86400, Condition= Not-Acknowledge
25 -0 After 24 hours Escalate alert to Sysadmin and Admin users, 1 day, 00:00:00 86400 Condition= Acknowledge.
After 24 hours I will acknowledge the alert. If it works I should only receive a reminder alert ever day until its resolved.

[Gsmith]

Unfortunately the above configuration did not work. I still received alert every hour after the action was acknowledged.
I had to create two action to get the notifications to work.

First action:
Step 1: Default 3600 (1hour) Not-ack for users //after 24 hours escalates to Site-Admin.
Step 2: Default 3600 (1hour) Not-ack Users and Site-admin

Second action using the same trigger:
Step 1: Default 84600 (24 hours) send alert site admin only when acknowledge but still a problem.

Is there a way to configure this all on the same action?

[Gsmith]

Started testing a new action with different configuration. Hoping to get what I need in one action instead of using two different ones.
Example;
Default operation step duration is 24 hour (84600 Sec)
Steps;
1-0 Alert Immediately Users. Duration 3600 sec, Conditions=Not-Acknowledge
25-0 after 24 hours Escalate alert to Sysadmin, 1 day, 00:00:00 86400, Condition= Not-Acknowledge.

[Gsmith]

Unfortunately this did not work as expected.
I received a notification immediately, then every hour Not-Ack.
After 24 hour it escalated and sent a notification to Users and admin. This was expected. Once it was acknowledge I still received a notification every hour to admin. I wanted it Only to send one notification every 24 hours to admin. I can make it work if there are two Actions made for one trigger but I'm unable to create one action to perform all the steps I need. Anyone have suggestion I could use on how to make this work?

[Gsmith]

Think I’m going insane, Help please.

A brief summery on my action/escalation that I’m trying to achieve;

Problem send immediately and every hour to users Not-Acknowledge. // No problems, this works as expected
Problem after 24 hours Not-Acknowledge send to Sysadmin. // No problems, this works as expected

If problem Acknowledge (i.e. problem still not resolved) send notification every 24 hours only to Sysadmin. // No joy I get a notification every hour to Sysadmin

My Action Configuration;

Default operation step duration = 3600s
1. Operation steps = 1-0 immediately duration=Default Send to all Admin Users
Operation Conditions = Event Acknowledged=Not Ack

2. Operation steps = 25-0 1Day, duration=86400 Send to Sysadmin
Operation Conditions = none
As you can see from my documentation in the Zabbix forum I’m unable to achieve this in one action, but I’m able to do this in two actions.

Steps work great until it’s acknowledge, but the problem is no matter how I configure the steps it still sends a notification every hour to sysadmin after its been acknowledge. I’m not sure if this is a bug.

Any help or direction would be appreciated

[duncan9562]

Can you screen shot your Operation details for step 1, and post them here?

[vso]

Think I’m going insane, Help please.

A brief summery on my action/escalation that I’m trying to achieve;

Problem send immediately and every hour to users Not-Acknowledge. // No problems, this works as expected
Problem after 24 hours Not-Acknowledge send to Sysadmin. // No problems, this works as expected

If problem Acknowledge (i.e. problem still not resolved) send notification every 24 hours only to Sysadmin. // No joy I get a notification every hour to Sysadmin

My Action Configuration;

Default operation step duration = 3600s
1. Operation steps = 1-0 immediately duration=Default Send to all Admin Users
Operation Conditions = Event Acknowledged=Not Ack

2. Operation steps = 25-0 1Day, duration=86400 Send to Sysadmin
Operation Conditions = none
As you can see from my documentation in the Zabbix forum I’m unable to achieve this in one action, but I’m able to do this in two actions.

Steps work great until it’s acknowledge, but the problem is no matter how I configure the steps it still sends a notification every hour to sysadmin after its been acknowledge. I’m not sure if this is a bug.

Any help or direction would be appreciated

What if you try steps 1-24 send to admin users
steps 25-0 send to sysadmin this way steps won't overlap and following rule will not get in your way:
Several operations can be assigned to the same step. If these operations have different step duration defined, the shortest one is taken into account and applied to the step.

[Gsmith]

VSO,
Thank you for the Reply, I'm setting it up as you suggested. Going to test it out in the next 24 hours.

[Gsmith]

Can you screen shot your Operation details for step 1, and post them here?

Hello duncan9562 Thank you for the reply, This was my old Config.

[Gsmith]

This is my new Configuration as VSO suggested

[Gsmith]

Limited step #1 to 24 did the trick. Not exactly what I want to happen for an Escalation, but it did solve one of my problems.

Thank you VSO for the suggestion. I guess if a trigger is going off Not-Ack for more then 24 hours I have a employee problem also.