Anyway to disable repeat smart alerts?

[dragon2611]

I have a drive in one of my proxmox boxes that has some bad sectors.

This drive is passed though to a VM and is part of a BTRFS Raid array and any data I care about is backed up elsewhere. I.e I should be able to quite comfortably have the drive or even the entire machine go bye bye and not suffer anything more than a bit of inconvenience since it doesn't host anything critical.

Now with that out of the way I'm not planning to replace a 3TB drive just because of 8 pending sectors considering it's been like that for about a month now and hasn't increased.

The following warning/error was logged by the smartd daemon:
Device: /dev/sdb [SAT], 8 Currently unreadable (pending) sectors

So is there anyway to stop it alerting me every day for this, but do it that in such a way that if the situation changes (i.e the number of pending sectors increases) then it will still alert for that.

 

[cpzengel]

Also interested without disabling smart itself. so can I seth threshold to 10 or higher?

 

[Stoiko Ivanov]

So is there anyway to stop it alerting me every day for this, but do it that in such a way that if the situation changes (i.e the number of pending sectors increases) then it will still alert for that.

AFAIK - not really with smartd (you can generally disable mail notifications, also for a single drive though).
A monitoring solution like CheckMK, or something more fancy usually has more options for those things.

 

[cpzengel]

I did not find any other ways to do.
We also check Proxmox fully with check_mk, a perfect Solution for that.
But check_mk also needs smartd running, so only Spamfilter will fix that.
We appreciate early Warnings for defective Sectors, but not every 24 hours.
The Question is how many uncorrectable Sectors a Device can Handle before you get a Data loss
ZFS is a good Soltion to prevent serious Problems

 

[Stoiko Ivanov]

* The check_mk module I used to use just collects the stats via smartctl (i.e. does not need to have smartd running).
* However I usually also run the regular tests via smartd, but have disabled the mail-feature of smartd (it should be disabled by removing the
` -M exec /usr/share/smartmontools/smartd-runner` in `/etc/smartd.conf` (check the script, and the plugins in /etc/smartmontools/run.d, for what else happens when it recognizes issues).