FreeNAS 9.3 NOT sending emails when alerts are present

[Stupendous Man!]

Little Back Ground: (Skip down to problem if you don't care)
Lets start real quick with my test bench. I'm evaluating FreeNAS using Virtual box. I want to play with it to see how it functions. One of the most important features of NAS software to me is the ability to get notification of problems and status. The email capability appears to be present, and would be great if it worked.

PROBLEM: Emails don't seem to be sending when Alerts are present. (Also there isn't much in the way of documentation of what gets emailed and when the emails are serviced. So I don't know what to expect). Reboots/Start ups aren't sending emails either. (I would think illegal shutdowns/restarts would be something that would send notification.

Settings/System Info:
- System->Information::Build = FreeNAS-9.3-STABLE-201509022158
- System->Advanced::Periodic Notification User = root (root's email is configured correctly otherwise the test would fail)
- System->Email:: FULLY Configured for usage of Gmail SMTP, with Successful test emails.

1. I was hoping that would get an email immediately after a powerup with Status even if it was good. (This Does NOT happen.
2. I currently "Plucked" a virtual HDD from one of my arrays to create an Alert, and I have yet to receive an email from the system.

If anybody can shed some light on this or inform me of a settings that I'm missing, that would be great.

I searched up and down forums and websites and I only came across one thread that seamed to be similar but it was from version 8.xx and the claims where that this was fixed in 9.xx

Thanks,

D

 

[joeschmuck]

Emails are not sent just because you shutdown and started FreeNAS. You will get one at 3AM indicating the system had rebooted. A VM doesn't work the same a a pure metal system does if you are not running pass through hard drives which is what you would really need to do. One thing for certain is do not make assumptions on how FreeNAS works, this is not your typical commercial product.

 

[Stupendous Man!]

Thanks for your response,

I think this clears things up a bit. I think without any code changes, a more comprehensive matrix of ALERTS, their Level (i.e. OK, Warning, Critical or whatever other levels you might have) and when/if they trigger an Email would be good for the documentation. Read below for my explanation/interpretation of the manual, and why having this matrix may clear up any misconceptions on how the system functions.

The lack of documentation regarding Alerts vs. emails is the problem. When and what triggers emails is somewhat unclear, and the periodic value is not well defined, "Nightly" is the only reference I could find.

I read the documentation and didn't get the result described. I don't think that the pass through HDD is my problem. I still believe an email should be sent based on how the system is responding in my Test setup.

I am able to simulate a critical error and FreeNAS correctly indicates a "CRITICAL" Alert. If the software recognizes there's an alert I would expect based on documentation that I would receive and email.

I found this regarding Email messages. From 22.Alert in the 9.3.1 Documentation:

Informational messages will have a green “OK” while messages requiring attention will be listed as a red “CRITICAL”. CRITICAL messages will also be emailed to the root user account. If you are aware of a critical condition but wish to remove the flashing alert until you deal with it, uncheck the box next to that message.

in the Email configuration section I have this:

Note: it is important to configure the system so that it can successfully send emails. An automatic script sends a nightly email to the root user account containing important information such as the health of the disks. Alert events are also emailed to the root user account.

Now, if you're saying that ALL Events, Critical Message(s), and/or Status Messages queue up and are only sent at 3:00 A.M. That's fine and basically answers my question. But this is certainly not clear.

Additionally:
I suppose the only assumption I made is that Critical messages based on the wording seemed to be separate from the "Nightly" emails. In the two instances of documentation it seems to imply that Critical Messages are emailed on their own event triggers.

I suppose I just figured if the Alert daemon is updating alerts every minute, and a Hard Drive is detected BAD or Missing that an email would be sent out immediately. Rather than waiting up to 23 Hours and 59 Minutes after the potential worse case scenario (HDD Failure at 3:01 A.M.)

 

[danb35]

No, critical alerts should generate immediate emails. @joeschmuck is correct that there isn't an email sent just because you booted your server (but it will result in system log entries that will be emailed as part of the daily security log), but if you pulled a drive, and the system noted that the drive is missing and thus gave a critical alert (flashing red warning in the GUI), it would ordinarily generate an email alert. As to why your system isn't, though, I don't have anything to suggest.

 

[Stupendous Man!]

@danb35 : Thanks, I think too much focus was on my mention of the boot/reboot email. I could careless if that comes in a status email nightly. The part that seemed wrong was that I have a critical Alert and no immediate email.

Now, the one thing that might be different is that my critical Alert was probably determined at startup, since the only way could produce a hard drive issue on an VM was to:
1. shut the VM down
2. Disconnect the Virtual Disks from the VM
3. then power it back up.

I wonder if the missing drive ALERT being generated at power up is somehow the cause of it not being emailed immediately.