Why when it goes below a threshold?  Why not simply remove partitions as space is needed dynamically?  Why pick an arbitrary number that is always going to be either too low or too high?  A correct number will never be picked.

Per offline discussion, sadly it is impossible to know when an app decided not to do something due to not having enough space.
Imagine the following scenario:
App X checks how much space is available, sees that it's below Y (internal to the app), decides to warn user.
Since the app didn't try to actually write, and the only system call it made was to check how much space is available, we are not able to detect what would be the right amount of space it needs.

Having a threshold of a GB today could solve just a shy of all possible interactions. For the very few remaining, we can have a policy.

Ping?

Hi all - this is a P1 bug - status update here?

Omri, can you help us define a spec about the policy which you mentioned in comment #2?

My assumption from Cyrus' request is that the policy would simply allow the admin to define a minimum threshold to trigger disk cleanup. Cyrus, does that answer your problem?

Weifang, since this is sys file related, can you lead this one? 

omrilio@ - Looped you into an offline discussion.

I just wanted to comment that I still think this is not addressing the real reason this issue was initially raised.  With that said, I'm all for having options, and having this option will in turn "fix" the problem I'm referring to because I'm assuming one will be able to just set the number to something outrageous and have it always do what they want.

What I mean by all of this, and what cyrusm sort of mentioned in comment 1, is people don't know what this new option should be set to.  Chromebooks come with different amounts of space.  It seems weird to have a policy that will intent to be global in nature (yes it can be applied to an OU but that seems even more absurd).

When I initially reported these types of issues in the other threads surrounding this topic, I asked there to be an option to remove ALL user profiles automatically.  Or maybe only keep 2?  e.g. As soon as someone logs off (cleanly) it would delete their profile.  In an environment with shared devices, there is never (rarely?) a need to keep user profiles on the devices.

The issues we face in our schools is exactly what omrilio said in comment 2-- Apps will CHECK to see how much space is available, and will simply cease to do their functions if too little is available (I'm looking at you screencastify).  So while this option will fix this issue because I'll just tell my devices to delete storage space when they have less than 10GB of space available, I can't see it really being useful for what is being discussed here because people aren't going to know how much or how little is safe to keep if they aren't simply trying to delete all of it like we are.

Weifang - ping here too.  Android apps require more space than Chrome apps so these bugs are becoming more important for us to fix please.  Thanks!

Our school board is also experiencing a lot of issues over the last couple weeks with storage on our devices being used up and not cleared automagically, resulting in frequent chromebook freezes and crashes.  I have been doing the time-consuming task of a USB recovery on each affected device that is reported to me (I am sure I only see a small fraction of affected devices), and I know that is only a band aid as it will recur once the storage becomes full again.  

I read in another post a solution that I think might be programmatically difficult to implement, but would seem to resolve most of our issues...I think it may have been overlooked as it was not presented clearly, so I will restate it here.

What if we were to implement a two-pronged solution where:
A) When an app requests the amount of available space on the device, the OS returns not the current free space, but the free space that WOULD be available if all other local profiles were deleted.
B) Perform disk cleanup automatically any time local storage gets low rather than on a timed event.  If we can pop up a GUI message stating storage is low when it gets low, why not rather delete old user profiles until storage isn't low any more...if you run out of profiles to delete, THEN pop up the GUI message.

Alexey - for your consideration. 

Assigning to atwilson@ for CrOS triage.

bartfab - I really think our team should own disk space cleanup. We've struggled with this behavior for years now.

Hello!
This bug is receiving this notice because there has been no acknowledgment of its existence in quite a bit of time
- If you are currently working on this bug, please provide an update.
- If you are currently affected by this bug, please update with your current symptoms and relevant logs.

If there has been no updates provided by EOD Wednesday, 12/19/18 (5pm EST), this bug will be archived and can be re-opened at any time deemed necessary.

Thank you!

Due to lack of action this bug has been Archived. If work is still being done on this issue or you are still experiencing this issue please feel free to re-open with the appropriate information.

My understanding is that we did significant work on disk cleanup in M-68 and we should be "just handling it" better now.

Can we clarify what this bug is tracking then? I *really* don't think we should have a "min free disk space" admin policy unless we really can't just handle low disk space gracefully on our own.