Regarding Alarms

[rajeevtomar]

It will be great if somebody could explain "Alarms". I could see lot of messages in Alarms section........these are becoming unmanegeable.

How to manage Alarms messages...???

Thanks, Rajeev

[Nate Bell]

Are you refering to the Alerts page (Actions tab)? At least, that's the closest thing to what you are talking about that I can find in my version of Zabbix. The alerts just show what actions were taken after a trigger gets tripped. So any time a type of media is used to send a message to a user, that gets logged in the Alerts/Actions page. This page shows when a message was sent, who it was sent to, the subject and the body of the message. If the status shows the message was sent, then all is well. If it say it was not sent, then look to the far right of the page for the error that was returned.

If you were talking about something else, then I hope you can glean something useful from this message.

Nate

[rajeevtomar]

Thanks for your reply.

I am using version 1.0. I am not taliking about Alerts. In my version of zabbix I can see Triggers, QUEUE, Alarms, Alerts etc.

I could get what Triggers, QUEUE and alerts are ..:-) But not what Alarms are.

I hope now you will be able to explain it.

Thanks, Rajeev

[Nate Bell]

Ah, I see. In alpha10 these are called Events (though the link goes to a file called latestalarms.php, so I think I'm speaking to you now).

Again, these are related to triggers. However, in this case they show when a Trigger is On or OFF. From the left to right, you have: when the state change of the trigger happened, the description you gave the trigger, what the state of the trigger is, and the severity level you set for the trigger. Of course, this may be different than version 1.0 (I can't remember), but I'll just keep talking.

When you define a trigger, Zabbix checks to see what its state is. If you jump over to the Alarms page, you'll see your new trigger with the status of UNKNOWN. After a bit, Zabbix will determine the status, and will either show ON or OFF. If something changes for the trigger, then the status will change, and will be printed on another line.

That's really all that is going on. The Alarms screen shows what state a trigger is in, and when that happened, and the Alerts screen shows what actions have been taken after a Trigger's status changes. If you have a few triggers that constantly change states, or even become unknown for a while, then you're going to get a lot of activity on this page.

Hope that helps,

Nate

[rajeevtomar]

Hello Nate,

Thanks for your reply.

You have explained most of the things in beautiful manner.

I have a few more questions to ask.

When I click on decription in alarms section. I see a Table in which I can see time, status, duration, sum and % columns.

Could you please explain what is the meaning of these columns, especially duration, sum and % columns....??

Please let me hear from you.

Thanks for your time.

Regards,
Rajeev

[nikoula]

Could you please explain what is the meaning of these columns, especially duration, sum and % columns....??

Hello,
I think this could help you.

Duration is the time that ran out for your item's state.
Sum is the sum of all durations for the same item's state for your host.
And % is corresponding to (duration/sum)*100 for the item's state.

Nicolas.

[rajeevtomar]

Hello Nicolas and Nate,

Can you please explain this with respect to my data of alarms...???

Samba's nmbd daemon is down on bwga092
{bwga092roc_cnt[nmbd].last(0)}<1

Time Status Duration Sum %
2005.Aug.26 16:04:56 FALSE 50.3 mins 50.3 mins 100%
2005.Aug.26 15:57:26 UNKNOWN 7.5 mins 7.5 mins 12.98%

I hope if you will explain this , then I would be ablle to understand it.

Please let me hear from you.

Thanks, Rajeev

[nikoula]

Samba's nmbd daemon is down on bwga092
{bwga092roc_cnt[nmbd].last(0)}<1

Time Status Duration Sum %
2005.Aug.26 16:04:56 FALSE 50.3 mins 50.3 mins 100%
2005.Aug.26 15:57:26 UNKNOWN 7.5 mins 7.5 mins 12.98%

ok,
as you can see you have 2 states for the trigger "{bwga092roc_cnt[nmbd].last(0)}<1"
duration's showing us that the unknown state lasted 7.5 minutes
and the false state (which is the last state here) lasted 50.3 mins

then sum shows us that the false state lasted 50.3 mins in total
if for example you have had another line with the false state during 5 mins sum would like 55.3mins in the earliest line of false state.

and for the % column i've forgotten exactly which ratio is

Nicolas.

[Nate Bell]

I just want to mention how funny those two posts were with all the stuck in the triggers and elsewhere.

Nate