Testing Parity Z Raid. Removed drive & no notifications?

[RichTJ99]

Hi,

I have a HP N40L microserver with 4x 2tb drives. I setup a zraid 1 when I set it up so it gave me about 6tb of storage. I then copied 5.95 tb of data to the drive (videos, etc) all copies of data that can safely be destroyed.

This morning, I shut down the server, removed one of the drives & booted it back up. I didnt get an email notification (which is setup), I logged into the webgui & it didnt mention that the raid array was missing a disk.

When I dug deeper into the view volumes, I do see that it says degraded. I just thought that Freenas would have notified me that I lost a drive. Is it supposed to do that?

I am planning on taking another 2TB drive, putting it in, rebuilding, then removing another drive & putting the original one I removed back in, merely to test how well its rebuilding, etc.

Thanks,
Rich

 

[cyberjock]

Sigh... another one of these threads.

Search for "zfsd" and you'll see responses to this question.

 

[RichTJ99]

Thanks, I guess the short answer is that Freenas doesnt do that by default, yet. It sounds like there is a script you can run to check the disk status. Is there a newb guide to doing it & getting it working since it is a common request?

 

[Dusan]

Hold it! You do not need zfsd to get warnings. zfs status -x will show DEGRADED status almost immediately. The FreeNAS alerter (alert.py) that drives the alert light in the GUI runs zfs status -x every 5 minutes. Also, the alerter is able and does send emails for CRITICAL messages (test it, remove root password and you will get a warning e-mail). However! I just discovered a bug in how alert.py parses the output of zfs status. The bug causes it to parse everything else than HEALTY as UNKNOWN. UNKNOWN is a WARNING and will only trigger a yellow alert light -- there is no email sent for WARNINGs.
The good news is that the fix is simple and William already merged my pull request (https://github.com/freenas/freenas/pull/30), so next FreeNAS will email you when your pool DEGRADEs.

 

[Dusan]

I tested the fix in a VM and a DEGRADED pool now triggers a red warning light + you receive a warning email within 5 minutes. :)

 

[RichTJ99]

Do you have any way I could add the fix into my freenas? If you don't mind spoon feeding me!

Or is there some sort of nightly build?

 

[survive]

Hi RichTJ99,

All you should have to do is edit the alerts.py file on your usb key.

Just mount the key read-write with "mountrw /" & edit /usr/local/www/freenasUI/tools/alert.py

You can find the exact line & what you need to add in Dusan's github commit above.

-Will

 

[RichTJ99]

So I just add ", re.M)"

Can I do this within the Webgui? If not, do I vi into the file itself?

 

[survive]

Hi RichTJ99,

That's what I did.....

-Will