Since I’m mentioning quick fixes to otherwise system-crippling errors, I should also make note of the “stale NFS file handle” errors that popped up after the power was accidentally cut in the middle of a large file transfer across my home network.
The resulting files were very similar to the busted ones that appear when I built systems in Crux and Arch, but started X without dbus and hal — undeletable, unremoveable, unmoveable, unrenameable, and exceptionally irritating.
Rather than eradicate an entire partition though, I got wise and tracked down a better solution: a filesystem check from a live environment. Forcing the check — because the drive is still marked clean, even though those files are real impediments — is necessary, and so for an ext2 drive, a simple e2fsck -f -v -p /dev/hdaX
did the trick.
Of course, after the filesystem was corrected, I had to re-rsync those missing files across the network, but that was only a minor inconvenience.
And … and … and that’s all. 😐
Pingback: hal’s day of reckoning « Motho ke motho ka botho