Pacman

Historical bug tracker for the Pacman package manager.

The pacman bug tracker has moved to gitlab:
https://gitlab.archlinux.org/pacman/pacman/-/issues

This tracker remains open for interaction with historical bugs during the transition period. Any new bugs reports will be closed without further action.
Tasklist

FS#33000 - [pacman] extend CheckSpace to check for available inodes

Attached to Project: Pacman
Opened by Christian Hesse (eworm) - Friday, 07 December 2012, 08:52 GMT
Last edited by Allan McRae (Allan) - Saturday, 09 February 2013, 03:51 GMT
Task Type Feature Request
Category Backend/Core
Status Closed
Assigned To No-one
Architecture All
Severity Low
Priority Normal
Reported Version git
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

Summary and Info:
Enabling CheckSpace makes pacman check for free blocks, but it does not notice if filesystems run out of free inodes. Would be great if pacman could check for free inodes as well.
This task depends upon

Closed by  Allan McRae (Allan)
Saturday, 09 February 2013, 03:51 GMT
Reason for closing:  Won't implement
Comment by Dan McGee (toofishes) - Friday, 07 December 2012, 18:52 GMT
f_ffree should be available everywhere f_bfree is, regardless of whether we are making statvfs(), statfs(), etc. calls.

However, this gets a little bit tricky. For instance, nilfs returns "0" for free inodes (but gives you a correct number for total); btrfs reports 0 for both numbers. I'd be wary of adding filesystem-specific code here; I think we'd just skip inode checks completely if we saw a 0 in either place?
Comment by Dan McGee (toofishes) - Saturday, 08 December 2012, 18:35 GMT
Thinking about this more, this seems a bit rediculous. What modern filesystem are you using that has such crazy inode limitations?
Comment by Christian Hesse (eworm) - Saturday, 08 December 2012, 19:09 GMT
The filesystem is ext4 with default settings. But probably you are right, there is just a small number of people ever hitting such a problem.

In my case I created Arch Linux live media, which has some directories split off into filesystems to share data for i686 and x86_64. When I hit the problem I tried to install a package with a lot of include files. Disk space was not a problem, but inodes were. (First I thought this is a problem of device mapper snapshots, took some time to find the real culprit.)

In the future I will switch to XFS which does not have an inode limitation. So I am fine if you close this with "won't fix", though I think this can be of real interest (for a small number of people).

Loading...