Well, now that I've experienced some sort of disk failure on my NAS, I'm sort of annoyed that I did not set up any sort of notification system.
This is just a standard Ubuntu box sitting in the office. It doesn't have any sort of mail or cloud integration setup.
What would you recommend I use for alerting?
Anyway, I suspect a drive controller failure. It connects but the data is garbage. Can't even get SMART out of it.
@janusfox We have our machines set up so they can send outgoing mail with sendmail, which means stuff like cron email Just Works™.
We use opensmtpd, and it's pretty simple to set up opensmtpd to relay through your usual mail server _and log in to send stuff_, but you don't really have to do that if you're okay with your mails looking like spam to the mail server.
Anyway, if you've got that working, you could just have a cron job that prints some output and dies if things are horribly wrong, and then you should get mail about it.
so in our /etc/smtpd.conf we've got (among other stuff)
janusfox 🍅
Frost, wolf of winter 🐺🎄
table remotepasswords file:/etc/smtpd/remotepasswords
action "relay" relay
action "relay-proxied" relay host "smtps://[email protected]" auth <remotepasswords>
match for local action "local"
match from any for domain "thunderjaw" action "local"
match from local for domain regex "^[^\.]+$" action "relay"
match from local for any action "relay-proxied"
(the "match from local for domain regex..." bit is probably so we can send mail to other LAN machines without it having to bounce through the mailserver in the sky)
(thunderjaw is this system's hostname)
and then in /etc/smtpd/remotepasswords
thunderjaw thunderjaw:hunter2
(the first "thunderjaw" here is just a local label. the "thunderjaw@" in the relay line looks it up by that. the "thunderjaw:" is the username used to log into the mail server.)