Alert Notifications (occasionally) Not Being Sent Via Smtp

[slibby]

We are running MONyog Ultimate 6.21, and occasionally our MONyog server fails to send an SMTP notification for warning- or critical-level alerts.

 

The affected monitors can be built-in, or custom monitors which we’ve created using CSOs.  

 

Everything is configured properly in the monitor itself, including the Value* and Alert condition javascript expressions (confirmed because the alerts work as intended sometimes, but not always).  The Notify? radio button is selected “Yes”.

 

The server(s) we are are monitoring are configured correctly for SMTP notifications, and the global SMTP configuration under Tools>Preferences>Mail settings are also setup correctly.

 

On the MONyog server itself, there are no relevant entries in the MONyog.log file.  We have even gone so far as to setup a “sanity check” custom monitor which does a simple “SELECT NOW();” against one of our servers every 12 hours, and it sends a Warning notification alert when the time is between 8:00AM and 5:00PM each day (currently doing the check at 12:54 and 00:54 each day), and then it resets back to OK status each night so the Warning alert will send again the next day.  However, even this alert is only being sent sporadically, even though the History for that alert clearly shows that it should have been sent every day since the initial setup of the monitor.

 

I have even tried scripting an API call to “_object=PrefMgr&_action=SmtpTest&from_name=[etc…]”, but this doesn’t work of course because the _object value is not “MONyogAPI”.

 

Is there anything else we can do to troubleshoot this issue or perhaps setup an automated daily test email from the server to ensure SMTP notifications are being sent as expected?

 

Thank you,

Steven Libby

[siddhant]

Hi,

 

If MONyog fails to send a notification alert it logs this error in the MONyog.log file. An example error which occurs due to incorrect credentials looks something like this:

smtpcallback.cpp(158) ErrCode:1 ErrMsg:Retrying PLAIN authentication after 535 5.7.8 http://support.google.com/mail/bin/answer.py?answer=14257 fp13sm181179



smtpcallback.cpp(158) ErrCode:2 ErrMsg:Can not authenticate to SMTP server: 535 5.7.8 http://support.google.com/mail/bin/answer.py?answer=14257 fp13sm181179

Please see if you get errors like this logged in your MONyog.log file. If this has anything to do with an error with your SMTP server, the error message can be found here.

 

Also. please check if you understand the concept of alert-able and have it configured to send alert notification immediately. Please follow these steps:

 

1. Click on the button in front of the server on the left panel in MONyog and select Edit server.
2. On the edit server page click on the Notification settings tab.
3. Check if the Send notifications over mail(SMTP)? is set to Yes. Also check if your mail settings are properly configured.
4. Now in the Send notification when alert-able set the value to 1.(default if 5)

The value in this field specifies the number of consecutive times for MONyog to encounter the same problem before sending the mail alert. In simple words, if set to 5 then MONyog needs to encounter the same problem for 5 consecutive times(or 5 consecutive data collections) before sending the mail alert. But setting it to 1 would ensure that the mail alert is fired the first time it encounters the problem. Please refer to the attached screenshot.

 

 

 

If this still doesn’t solve the issue, please share with us the events.data file for the problematic server in the MONyog’s data directory. Upload it to Dropbox and share with us the link for the same by writing us back to [email protected]

[slibby]

Hello Siddhant,

 

We are still not showing any relevant entries in MONyog.log around the time frames when these alerts should have been sent.  Also, the “alert-able” configuration has been set to 1 instead of the default value of 5 for all servers since prior to our realization that the alerts were not being sent occasionally.

 

I am sending an email with the Dropbox links to the events.data files for two of our servers which have been configured to alert on our ‘Notification Test’ monitor.  The CSO was originally applied to one server, then we switched it to apply to a different server early this morning.

 

Thank you,

Steven Libby

[Author]

Posts