[bug] Multiple simultanious triggers and not enough mail to match

[ReefShark]

Here's a problem I've noticed in most Zabbix (1.1bx) versions I've run so far; when a number of triggers occur at the same time, the email for every single trigger isn't send out.

Here's how I noticed;
I have a number of linux machines and logging on to every single one and recompiling Zabbix whenever we get a version bump takes a lot of time. This and a few other simular tasks are best done by scripts remotely executed through ssh. So after an afternoon of scripting, I can update every Zabbix installation on every (linux) server. Today I upgraded to 1.1beta7 and within an hour, it was compiled on all linux hosts. Zabbix monitors the agent version and checks it's status every 30 minutes. You can imagine I get about 10 to 15 triggers at once when I do a version bump this way. But I only receive an email for 2 or 3 of those triggers. I definately want the other mails too. In this case, it's something I know and caused myself, but if a router fails and a dozen hosts become unreachable, I want to know about it, even if I don't have my tr_status.php page within view.

[ReefShark]

Another thing I noticed about 1.1beta7 is the abundance of email; the server just lost connection with a server (for a minute) and responded by sending me 4 emails about it... and only the trigger OFF email, not the trigger ON email.

[bbrendon]

I can't confirm this but will keep an eye out. Someone has 1200 hosts monitored with zabbix. Maybe he can chime in... i forgot who it was though

You might want to check your suckers, etc and make sure the server is handling the load okay. Assuming its not a bug, thats the only solution i can think of for you, other than you may have a configuration error.

[bbrendon]

Another thing I noticed about 1.1beta7 is the abundance of email; the server just lost connection with a server (for a minute) and responded by sending me 4 emails about it... and only the trigger OFF email, not the trigger ON email.

My system only sends ON notifications and I've never gotten 4 emails about it. Maybe test it that way?

[markpr]

I have held off reporting my trigger notifications problems until now.. (1.1b5 and 1.1b6) because I thought it just must be me...!

I have 2 hosts both with identical triggers. The trigger they are all using is: {myhostname:icmpping.max(900)}=0} where my "myhostname" is the host name of the host being monitored

They are both a member of the same group and share the same action. Both hosts have events showing the change of state between ON and OFF. One host only ever sends an action with an ON status, the other host sends an action for both ON & OFF. I can't see any difference between the two hosts.

Mark

[bbrendon]

Sounds like this problem might only exist when using the OFF functionality?

[markpr]

Are you also only seeing ON notifications or have you configured it this way (I know in the Alphas you were able to specify these independently - ON|OFF or BOTH I recall)? I do get some ON & OFF notifications from some hosts but the majority are only sending out ON notifications. I should have also added that all hosts are sending notifications to the same user/media.

[bbrendon]

Are you also only seeing ON notifications or have you configured it this way (I know in the Alphas you were able to specify these independently - ON|OFF or BOTH I recall)? I do get some ON & OFF notifications from some hosts but the majority are only sending out ON notifications. I should have also added that all hosts are sending notifications to the same user/media.

I only have ON configured.

[heatherf03]

I am having similar issue with 1.1beta8 server. Below is the email I sent to my co-worker who is the Zabbix Admin and he suggested I post a question to this forum. Anyone find a solution to this, yet?


I created the "TeleTechUNIX" template and defined triggers for it. I linked each host that is in the "NISC - Telecom" and "NISC - Telecom Test" host groups to the template "TeleTech UNIX". I verified each host's agent.ping item is collecting and triggering to "on" when I adjust the value to "#2". I receive the number of email notifications that I define in "Actions" where "Send message to" is "Tele Tech" for three of the servers (02520dbs, teconv, tpro), but not the other three (tedbs01, telecd, bill). I created a trigger just for bill for the agent.ping and I received the email notifications. Do you know why when I trigger the template to send emails for all the servers for agent.ping that I am only receiviing for three servers? I wonder if there is a limitation on the server that is only allowing three servers at a time? Any other ideas?

Server, TeleTechUNIX, is not responding {TeleTechUNIX:agent.ping.avg(1200)}#2


I also triggered to "on" two triggers for bill and received three email notifications for the free space is low on /pro, but no emails for the agent.ping. This leads me to believe there is some type of limitations on the email notifications, unless I have something configured incorrectly for what I'm trying to accomplish.

[ReefShark]

Mass-installing beta8 yesterday (half an hour, almost 20 minutes, bash scripts and ssh for the win) allowed me to recheck the problem with the new release. I got an email for every server, none missing. So that seems to go better.

I do get 3 emails when things go wrong (certain triggers fireing). My actions are:
All Zabbix admins (me):

Trigger severity = "Warning"
Trigger severity = "Information"
Trigger severity = "Not classified"

The entire ICT Services department (including me):

Trigger severity >= "Average"

I just got a few mails from a windows host (severity average) and 3 mails; two adressed to me, one adressed to the general department email adress.

Either Zabbix is a bit wacky or I really screwed up my config.