On Tue, 8 Mar 2022 07:43:27 +0000 Andrea Verni Andrea.Verni@u-blox.com wrote:
I'm running NAV 5.3 and have issue with alertengine and email notifications. Email seems to remain stuck in the queue until I force a reboot of the VM.
Rebooting sounds pretty drastic and unnecessary for anything that doesn't involve a kernel upgrade; how about just restarting the NAV process that has issues? `nav restart alertengine`
The error message I see in /var/log/nav/alertengine.log is: 2022-03-08 08:35:10,707 [WARNING] [nav.alertengine.alertaddress.send] Not sending alert 5824 ****** as handler Email is blacklisted: [Errno 110] Connection timed out 2022-03-08 08:35:10,716 [WARNING] [nav.alertengine.alertaddress.send] Not sending alert 5825 to ****** as handler Email is blacklisted: [Errno 110] Connection timed out 2022-03-08 08:35:10,720 [INFO] [nav.alertengine.check_alerts] 0 new alert(s), sent 0 alert(s), 910 queued alert(s), 0 alert(s) deleted, 910 failed send(s), 0 ignored
However, by running a tcpdump I do not see any outbound/inbound connection to the configured email server.
That's because the email handler has been blacklisted by alertengine.
A handler that keeps producing errors will be "blacklisted" internally by the alertengine process. This means that alertengine will stop using it for the rest of the process runtime.
In this case, the email handler has likely been timing out while trying to talk to your SMTP server, eventually causing it to be blacklisted by alertengine - which stops any further attempt at sending e-mail for the lifetime of that process.
A simple restart of the process will kick things back into gear.