Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

flushchanges: corrupts gentoo/pentoo repos #305

Open
GoogleCodeExporter opened this issue Aug 13, 2015 · 4 comments
Open

flushchanges: corrupts gentoo/pentoo repos #305

GoogleCodeExporter opened this issue Aug 13, 2015 · 4 comments

Comments

@GoogleCodeExporter
Copy link

I recently learned that deleted files are ignored when saving changes with 
flushchanges (tnx ZC) because of aufs/squashfs.

This effectively corrupts the local gentoo/pentoo repositories.
Example:
1. Boot into virgin LiveCD
2. run 'layman -S', then 'flushchanges'
3. reboot the LiveCD
4. run 'emerge -p pentoo-installer'
==> Error: masked by: corruption
Full output see attached file "emerge-pentoo-installer.error.log"

Reason for the error:
Files being in the repo but not in the Manifest of the ebuild.
In this case "pentoo-installer-20131013.ebuild"

I have a hacky workaround, ....

Original issue reported on code.google.com by [email protected] on 27 Oct 2014 at 5:49

Attachments:

@GoogleCodeExporter
Copy link
Author

Here's the workaround:
1. Boot the virgin LiveCD (flushchanges never used on it)
2. sync the repos with 'emerge --sync' and 'layman -S'
3. Then run the attached script 'recreate_repo_lzm_files'
4. Then delete these files from '/mnt/cdrom/modules/':
- pentoo_overlay-20140922.lzm
- portage-20140922.lzm
5. Reboot! The deleted files were mounted!

Explanation of the script:
It creates new squashFS files for these 3 folders:
- /usr/portage
- /var/cache/edb
- /var/lib/layman

Original comment by [email protected] on 27 Oct 2014 at 5:56

Attachments:

@GoogleCodeExporter
Copy link
Author

Yes, it IS hacky, but it seems to work here.

The tricky part was getting the paths right when creating the new squashFS 
files, so maybe this comes handy one day.

Original comment by [email protected] on 27 Oct 2014 at 5:58

@GoogleCodeExporter
Copy link
Author

this will be accepted when the aufs modules handle whiteout properly.  I don't 
like this hack, or the fact that there will be dozens more caused by us not 
handling whitespace.

then again, use persistence properly and it should actually work fine.

Original comment by sidhayn on 27 Oct 2014 at 7:25

@GoogleCodeExporter
Copy link
Author

Original comment by blshkv on 28 Oct 2014 at 12:32

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant