[V.Low priority: Possible bug:] Yellow alert drive free space, 81% consumed

[diskdiddler]

Hi all,

I reported on this 6 to 8 weeks ago, maybe even 12, people explained to me why it's fine and I need not panic.
The thing is, in the time since I first got my 80% full alert in the FN UI, the figure has CONSISTENTLY remained at "81% in use" yet I know I've actually added hundreds of GB of data, I suspect I might be at 82/83/84 - maybe even 85% full.

Does this "81%" in the alert get re-calculated for the alert? I've never unticked it, to remind me to keep trying to free up more space (and yes, I do eventually intend to)

This is a very small issue, I could even be wrong but I'm pretty sure several hundred gb, should have infact ticked it up to higher than 81%

 

[dlavigne]

It will stay as a yellow alert as long as it is over 80% and will change to a red alert when you hit 90%.

 

[diskdiddler]

That's not the issue, the issue is the % is sitting solid at 81% despite hundreds of gb's being written to the system. I believe the 81% was first calculated and text inserted into an alert but the actual % amount is not being re-calculated.

 

[diedrichg]

Since it's a persistent nag then if I were you I'd go ahead and check it off as acknowledged. I'm betting that the alert is dynamic but not until you acknowledge it and until then it is static at the capacity when the alert was first triggered. Upon the next trigger I assume it announce the new capacity.
(this is all personal speculation though and not from experience)

 

[diskdiddler]

I believe the 81% was first calculated and text inserted into an alert but the actual % amount is not being re-calculated.

I have no issue with the nag, I have issue with it not re-calculating the figure.

 

[diedrichg]

And I bet it WILL recalculate once acknowledged. I imagine it like a stopwatch; if you press the split button you freeze that calculation but as soon as you release the split then you get to see the running time. Again, speculation and no facts...

 

[diskdiddler]

What?
If I acknowledge the alert, I will no longer have the alert? I like the idea of the alert to keep a constant reminder to the administrator "hey you need to fix this issue"

Are you saying that if I acknowledge the alert, it will re-calculate and just come back?

 

[BigDave]

:DThink of it like an idiot light on the dashboard of a car, don't worry about how many quarts of oil you are down,
just pop the hood and fix it ;)

 

[Bidule0hm]

The alert will stay in the list but the icon would be back to green.

 

[rogerh]

You know that the percentage fullness is readily visible in the GUI on the first tab (Volumes) under Storage? As others have said, the alert is a one-off thing, not a continuous calculation. It may be best to untick it, otherwise you may miss a new alert about something else because the alert button will already be red.

 

[cyberjock]

From what I've seen it updates itself at regular intervals. What those intervals are is anyone's guess. It also depends on what you are using the pool for. If you created a zvol that is 81% of the pool and you do NOT check the sparse volume it will immediately consume that much space. You can then fill the iSCSI drive with all the data you want and it will stay at 81%. Why? Because the zvol is already taking up the space.

There are plenty of theoretical scenarios where you can write TB of data to a pool and that number will never change. Even your darn browser cache can bone you. But earlier today I saw someone that was at 97% and then later 98%. So it *definitely* updates.

 

[rogerh]

When you say it updates itself, is the alert message edited, or is a new alert generated when a new milestone is reached?

 

[cyberjock]

It said 97%, and later we went to the WebGUI and it was 98%. We didn't check the checkbox or anything. The "disk full" message was simply updated. I believe those messages are regenerated every so many minutes (5?) so if the error condition were to change from 97% to 98% I'd expect that it would simply provide the new value at that time. By the same token, if you delete a bunch of files and go below the alarm, it doesn't instantly clear the alarm either.

 

[rogerh]

Thanks, that clarifies things.

 

[diskdiddler]

Full disclosure / credit: I was incorrect about this issue. Mine has finally updated to 83% full. So clearly it's actually working.