[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-03-28 Thread Launchpad Bug Tracker
This bug was fixed in the package systemd - 255.4-1ubuntu5 --- systemd (255.4-1ubuntu5) noble; urgency=medium * No-change rebuild against libcurl4t64 -- Steve Langasek Sat, 16 Mar 2024 07:04:30 + ** Changed in: systemd (Ubuntu Noble) Status: Fix Committed => Fix

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-29 Thread Nick Rosbrook
I have uploaded a change to add the 30d cleanup age. ** Changed in: systemd (Ubuntu Noble) Status: Triaged => Fix Committed -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is subscribed to systemd in Ubuntu.

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-29 Thread Nick Rosbrook
** Changed in: systemd (Ubuntu Noble) Status: Confirmed => Triaged ** Changed in: systemd (Ubuntu Noble) Assignee: (unassigned) => Nick Rosbrook (enr0n) -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is subscribed to systemd in

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-29 Thread Julian Andres Klode
I agree with Nick, regular cleaning *and* cleaning at /boot is best behavior. ** Tags removed: rls-nn-incoming ** Tags added: foundations-todo ** Also affects: systemd (Ubuntu Noble) Importance: Wishlist Status: Confirmed -- You received this bug notification because you are a member

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-23 Thread Nick Rosbrook
** Tags added: rls-nn-incoming -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is subscribed to systemd in Ubuntu. https://bugs.launchpad.net/bugs/2019026 Title: systemd /tmp cleaning is suboptimal Status in systemd package in Ubuntu:

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-19 Thread Paride Legovini
Looks like the experiment worked: having e /tmp 1777 root root 30d deleted a test directory which has mtime and atime >30d old (done via: touch -d '29 days ago', and then left to age over the weekend). -- You received this bug notification because you are a member of Ubuntu Touch seeded

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-16 Thread Brian Murray
It looks like something on our autopkgtest-cloud-worker may also be regularly changing the contents of /tmp but we've set up an experiment for over the weekend and will report back. -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-16 Thread Paride Legovini
AIUI he suggests keeping the boot time cleanup, and that it could be nice to have the 30d behavior back. In any case that's what I'd prefer. I know we're speaking of file age; certainly not of blindly deleting everything every month or so :-) -- You received this bug notification because you are

Re: [Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-16 Thread Steve Langasek
On Fri, Feb 16, 2024 at 01:06:07PM -, Paride Legovini wrote: > Maybe we could make that 40d, as 30d is likely to be a time interval at > which a lot of periodic things happen (e.g. an off-site backup). The period here is the age at which files are considered old and to be clean up, not the

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-16 Thread Steve Langasek
You say "Nick's preference", but it's Nick who took the position here that the default behavior should not change? -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is subscribed to systemd in Ubuntu. https://bugs.launchpad.net/bugs/2019026

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-16 Thread Paride Legovini
FWIW I agree with Nick's preference (clean at boot && clean files older than 30d). Maybe we could make that 40d, as 30d is likely to be a time interval at which a lot of periodic things happen (e.g. an off-site backup). A retention period >30d is less likely be synchronized with it in an unlucky

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2024-02-14 Thread Brian Murray
The Ubuntu QA team encountered an issue where our autopkgtest-cloud- workers in the prod-proposed-migration environment ran out of free space in /tmp because these are production servers which are rebooted very infrequently. Due to some bug in the autopkgtest-cloud or autopkgtest code there were

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2023-06-12 Thread Nick Rosbrook
> how would it impact you if it was NOT cleaned at reboot? I just like it. It's easier for me to keep track in my head that "oh, this file in /tmp will only be around until I reboot", as opposed to trying to keep track of the actual age of a file. To be clear, I am not suggesting that my use case

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2023-06-09 Thread Steve Langasek
(setting bug back to New because I don't see any request for information) ** Changed in: systemd (Ubuntu) Status: Incomplete => New -- You received this bug notification because you are a member of Ubuntu Touch seeded packages, which is subscribed to systemd in Ubuntu.

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2023-06-09 Thread Steve Langasek
The reason it's specifically important to me not to clean at boot is that schroot bind mounts /tmp by default but does not bind mount /var/tmp by default, so I am accustomed (since long before the systemd behavior became the norm) to using this directory for sharing data between the host system

[Touch-packages] [Bug 2019026] Re: systemd /tmp cleaning is suboptimal

2023-06-09 Thread Nick Rosbrook
I personally have become very accustomed to the current behavior where /tmp is emptied on reboot -- I have no idea what most users would say about this, so I wonder if changing that behavior would be unwelcome. I do think it could be nice to have the 30d behavior back, however. In other words, if