NEMS Linux Community Forum

First... I don't believe this is bug... at all.

I finally bit the bullet and upgraded to 1.3 a few weeks ago. But over time had noticed that notifications weren't working. So yesterday I backed up the config again and reloaded from scratch then restored the backup.nems file again. I then disconnected a monitored host and waited.

It looks to me as though my restored config is stepping on something, causing the device to try and use a mail agent that doesn't exist in the new version. Remember that early on I had installed Postfix on the earlier verion to get notifications to work. But this time I decided to not mess with the device at all but to simply use it as it was intended. :-)

What I'm seeing in the syslog file is this:

/bin/mail -s "** PROBLEM Host Alert: Node04 is DOWN **"

Not sure where to go from here as I really do not want to modify anything at all this time.

Looks like you modified the email notification commands, am I right?

Best thing for you to do in this case (so I can help) is to email me your backup.nems file to review what's actually happening here. Please send to [email protected]

Thanks.

Real early on yeah... I had installed Postfix as a local mail server to get notifications working. If I remember right it was back at v 1.1 or something. Subsequent updates I just kept it that way. But going to the latest I decided "No. DON'T DO that. Use it the way it was INTENDED." :-D

I'll take a crack at sending you the earliest backup I used to go to the 1.3 version but it's a 36MB file and I don't think either of my two mail accounts will allow it. If not... I'll just reload the SD card again, and then not restore but rather add the devices to monitor as if it was the first time NEMS for Pi had been deployed to the network. I'm not really all that concerned about losing any of the 'history' really since the network it runs on is mostly just a play thing for me. ;-)

Yeah... can't do it. Google bounced it. Oh well... it'll be a lot easier to just start from scratch and add the target hosts to monitor as an all new installation.

No worries. :-)

OK. I did a totally new installation. Didn't restore any backup and just started from scratch. Only modifications were to change the port that sshd is listening on (I pretty much never use the default) and added snmpd so the other Pi I have acting as a Cacti server can watch it. I configured it to use the internal Exchange Server as the smtp server, set the "From" address to be "nems@<my domain.com>" as it should be. Mail should be going to "administrator@<mydomain.com>" but I'm not getting anything. Noticed this morning in the Exchange Server's Application Event Log that an attempt was made by the device to send an e-mail in but Exchange saw it as a looping condition:

A non-delivery report with a status code of 5.3.5 was generated for recipient rfc822;nagios@localhost (Message-ID  <505709.030755924-sendEmail@nems>).

The Exchange Server is configured to accept incoming mail from all of the devices on the LAN and has been doing so for several years (not a good practice really, but still) but for some reason that I can't figure out it really doesn't like this Pi... LOL

OMG I'm an IDIOT!!!!

I was NOT paying attention in NConf... the e-mail address of the 'administrator'... was still set to the default.

<walks away... hanging his head in shame...>