Bug#754340: Unable to run fsck manually when instructed to do so
Am 13.07.2014 22:17, schrieb Bas Wijnen: On Sat, Jul 12, 2014 at 12:59:04AM +0200, Michael Biebl wrote: Am 12.07.2014 00:34, schrieb Bas Wijnen: When fsck failed with this message before, I could do: mount / -o remount,ro fsck / Now, and I'm guessing this is a change on the part of systemd, that first command (remount read-only) fails with the message that the file system is busy. Having no bootable computer and thus no internet, I was unable to figure out what was keeping it busy, and how I was supposed to stop it. This is the information that I think should be part of the please run fsck manually-message, because that won't work without it. I think there is no general answer to that. There most likely simply was a process keeping your (root) fs busy. So I would have tried stopping one service after another. I didn't start any processes. The problem happened in fsck, at which point no process should be allowed to write to the file system (except fsck itself). After failing, it gave me a rescue shell with which I cannot remount the fs read-only. I think whatever is keeping it busy must have been started just before spawning that shell? Or is it the shell itself? I'd be fine with stopping all services, but I'm not familiar with how to do that either. If this is the solution, please add that instruction to the message. But would services which prevent the disk from being remounted read-only be started before fsck is finished? The rescue shell is part of fsck, I think(?), so nothing should have been started after it failed. Since we don't know what kept your fs busy, I'm at a loss how we are supposed to write instructions/documentation. Could you boot into single user mode (via the single kernel command line option) and test if you can remount / ro. This rescue mode should be similar / the same what you get when fsck fails. If your fs is busy, start with dumping the ps output and the systemctl output, so we can which processes / services are running. -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature ___ Pkg-systemd-maintainers mailing list Pkg-systemd-maintainers@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-systemd-maintainers
Bug#754340: Unable to run fsck manually when instructed to do so
On Sat, Jul 12, 2014 at 12:59:04AM +0200, Michael Biebl wrote: Am 12.07.2014 00:34, schrieb Bas Wijnen: When fsck failed with this message before, I could do: mount / -o remount,ro fsck / Now, and I'm guessing this is a change on the part of systemd, that first command (remount read-only) fails with the message that the file system is busy. Having no bootable computer and thus no internet, I was unable to figure out what was keeping it busy, and how I was supposed to stop it. This is the information that I think should be part of the please run fsck manually-message, because that won't work without it. I think there is no general answer to that. There most likely simply was a process keeping your (root) fs busy. So I would have tried stopping one service after another. I didn't start any processes. The problem happened in fsck, at which point no process should be allowed to write to the file system (except fsck itself). After failing, it gave me a rescue shell with which I cannot remount the fs read-only. I think whatever is keeping it busy must have been started just before spawning that shell? Or is it the shell itself? I'd be fine with stopping all services, but I'm not familiar with how to do that either. If this is the solution, please add that instruction to the message. But would services which prevent the disk from being remounted read-only be started before fsck is finished? The rescue shell is part of fsck, I think(?), so nothing should have been started after it failed. Do you have a /var/log/journal/ directory and is this on the root directory? Might be that journald kept your fs busy. No, I don't have that. However, I just noticed that it is on LVM, currently running with one physical volume on one disk. I don't think that should prevent remounting it read-only though... Thanks, Bas signature.asc Description: Digital signature ___ Pkg-systemd-maintainers mailing list Pkg-systemd-maintainers@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-systemd-maintainers
Bug#754340: Unable to run fsck manually when instructed to do so
Am 12.07.2014 00:34, schrieb Bas Wijnen: When fsck failed with this message before, I could do: mount / -o remount,ro fsck / Now, and I'm guessing this is a change on the part of systemd, that first command (remount read-only) fails with the message that the file system is busy. Having no bootable computer and thus no internet, I was unable to figure out what was keeping it busy, and how I was supposed to stop it. This is the information that I think should be part of the please run fsck manually-message, because that won't work without it. I think there is no general answer to that. There most likely simply was a process keeping your (root) fs busy. So I would have tried stopping one service after another. Do you have a /var/log/journal/ directory and is this on the root directory? Might be that journald kept your fs busy. Michael -- Why is it that all of the instruments seeking intelligent life in the universe are pointed away from Earth? signature.asc Description: OpenPGP digital signature ___ Pkg-systemd-maintainers mailing list Pkg-systemd-maintainers@lists.alioth.debian.org http://lists.alioth.debian.org/cgi-bin/mailman/listinfo/pkg-systemd-maintainers
