Email not beeing sent

[FourierXform]

I have searched through the forums, however none of the solutions provided helped.

I have done the following:

Clicked "CONFIG":

Description: Email Alert
Type: mail
SMTP server: smtp.myserver.com
SMTP helo: myserver.com
SMTP email: [email protected]

I use Sendmail and checked to make sure I could send mail to the [email protected] user.

Clicked "USERS"

Under "Configuration of Users", I have my username listed and then...

Clicked "Media"

Type: Email Alert
Send to: [email protected]
Use if severity: (I highlighted all of them)
Status: Enabled

I set up the following for a trigger as a test:

Clicked "TRIGGER":

Description: Test Trigger - Telnet
Expression: {server.myserver.com:net[listen_23].last(0)}=0
Severity: High
URL: server.myserver.com:23

Clicked "ACTIONS" next to "Test Trigger - Telnet":

Send Message To: Single User
User: myusername
When Trigger Becomes: On or Off
Delay between messages (in sec)
Subject: Test Trigger - Telnet
Message: Message ...
Scope: This trigger only

When I checked through the mail log, I see the following:

Aug 11 17:21:25 server sm-mta[2127]: j7BLLPCZ002127: localhost [127.0.0.1] did not issue MAIL/EXPN/VRFY/ETRN during connection to MTA

I am not sure what is going on. Any help would be much appreciated.

Thanks!

[James Wells]

Greetings,

SMTP server: smtp.myserver.com

First question is 'Can the Zabbix server resolve and reach smtp.myserver.com?' You can verify this by checking the sendmail logs to see if the zabbix server is attempting to send email.

Description: Test Trigger - Telnet
Expression: {server.myserver.com:net[listen_23].last(0)}=0
Severity: High
URL: server.myserver.com:23

Clicked "ACTIONS" next to "Test Trigger - Telnet":

Based on what you are showing here, the trigger will not change states until telnet ceases listening to port 23. You will want to turn telnet on and off a couple times, every few seconds, while watching the sendmail logs on the server smtp.myserver.com.

Aug 11 17:21:25 server sm-mta[2127]: j7BLLPCZ002127: localhost [127.0.0.1] did not issue MAIL/EXPN/VRFY/ETRN during connection to MTA

Right off the bat, I would guesstimate that you have configured a zabbix item to monitor port 25 and that is what this is showing, not that it is attempting to send email. Additionally, as you configured the SMTP server to use smtp.myserver.com, you wouldn't see 127.0.0.1 here, you would see the IP or the name of the zabbix server here if it were ttempting to send email.

[FourierXform]

Greetings,

Greetings to you as well and thank you for taking time to answer my questions .

First question is 'Can the Zabbix server resolve and reach smtp.myserver.com?' You can verify this by checking the sendmail logs to see if the zabbix server is attempting to send email.

Yes. In fact the Zabbix server resides on the same machine as the smtp server (Sendmail).

Based on what you are showing here, the trigger will not change states until telnet ceases listening to port 23. You will want to turn telnet on and off a couple times, every few seconds, while watching the sendmail logs on the server smtp.myserver.com.

I don't have a Telnet server running at all. If I look in the "TRIGGERS" section right above the "USERS" link, it does indeed show that the trigger has been set off.

Right off the bat, I would guesstimate that you have configured a zabbix item to monitor port 25 and that is what this is showing, not that it is attempting to send email. Additionally, as you configured the SMTP server to use smtp.myserver.com, you wouldn't see 127.0.0.1 here, you would see the IP or the name of the zabbix server here if it were ttempting to send email.

Got ya!

Unfortunately I am still not receiveing any emails nor is anything beeing shown in the mailog that would indicate an attempt was even beeing made. There is nothing in the Zabbix logs either. Any other suggestions?

Thanks!!

[James Wells]

Yes. In fact the Zabbix server resides on the same machine as the smtp server (Sendmail).

I apologize if this question may sound stupid, based on your responses, but... On the server, when you do a 'host smtp.myserver.com' (The host name you are using in the SMTP Server option), it returns the IP address of the local machine?

I have to ask this because, Zabbix does use normal name service look ups to find the host to send the mail to. And what you are describing indicates that Zabbix is unable to find the IP of the server listed in the SMTP Server option.

Based on what you are showing here, the trigger will not change states until telnet ceases listening to port 23. You will want to turn telnet on and off a couple times, every few seconds, while watching the sendmail logs on the server smtp.myserver.com.

I don't have a Telnet server running at all. If I look in the "TRIGGERS" section right above the "USERS" link, it does indeed show that the trigger has been set off.

I guess I wasn't real clear here. By the time you created the trigger, the first check for it had already been performed, as such there was no change to state. As soon as you turned on the trigger, it went into a failed state, but since it was already in a failed state there was no change here as well. All well and good so far, however, this is where things break up a bit. Since the trigger is already in a failed state, no action will be taken when you turn the action on, instead, the action will not be taken until such time as the trigger changes state again.

Hopefully that clears things up a bit.

[FourierXform]

I apologize if this question may sound stupid, based on your responses, but... On the server, when you do a 'host smtp.myserver.com' (The host name you are using in the SMTP Server option), it returns the IP address of the local machine?

I have to ask this because, Zabbix does use normal name service look ups to find the host to send the mail to. And what you are describing indicates that Zabbix is unable to find the IP of the server listed in the SMTP Server option.


I guess I wasn't real clear here. By the time you created the trigger, the first check for it had already been performed, as such there was no change to state. As soon as you turned on the trigger, it went into a failed state, but since it was already in a failed state there was no change here as well. All well and good so far, however, this is where things break up a bit. Since the trigger is already in a failed state, no action will be taken when you turn the action on, instead, the action will not be taken until such time as the trigger changes state again.

Hopefully that clears things up a bit.

Hi James,

I think you hit it right on the head about the state chage of the sevice beeing monitored. I actually did get the emails to work as last night I was emailed a message saying that the processor load was above 5. However, that doesn not mean anything to me. Shouldn't the processor load be in some sort of decimal or percentage form? I guess I will have to research this. In any case, thank you very much for your time and help!

[James Wells]

Excellent!!! Glad, I was able to help.

Hi James,

I think you hit it right on the head about the state chage of the sevice beeing monitored. I actually did get the emails to work as last night I was emailed a message saying that the processor load was above 5. However, that doesn not mean anything to me. Shouldn't the processor load be in some sort of decimal or percentage form? I guess I will have to research this. In any case, thank you very much for your time and help!

As for the load email, not sure how much you understand the load value on a system, but in Unix, there are three different load values for the system.

When you do a 'who' or 'w' command on a Unix system, you will see the load average currently, over the last 5 minutes, and over the last 15 minutes. What this is actually indicating is the number of processes waiting for CPU time. So the email your server has 5 or more processes waiting for CPU time.

The current load average isn't as big a deal as the 5 minute and minute minute values. Current load average of up to about 8, for a single CPU, is fairly normal at different times of the day, while 5 minute load averages of 5 or so, on a single CPU, indicates that your machine is getting a decent workout.

I hope that clears up the alert email you recieved.