On Thu, 31 Dec 2015 13:39:39 -0800 Russell Senior <[email protected]> dijo:
>>> The device /dev/sr0 is typically the first CD/DVD device found by >>> the operating system (and /dev/sr1 would be the second). All it's >>> saying (loosely) is that the device wasn't cleanly closed before >>> eject and an inode is still being "accessed" even if no available >>> any longer by some process. >>> >>> This is why you can't mount anything new at this point with the >>> device still being "busy", according to the kernel. > >John> Is there any way to fix it short of rebooting, which is a PITA? > >How are you removing the disk? Do you "unmount" it before pushing the >button on the disk drive? In the past I always ejected the media from the optical drive either by right-clicking on the drive as listed in Thunar "Places" and then Eject, or by physically pushing the button on the drive. Both always ejected the media, but apparently once recently the drive wasn't ejected cleanly. I have never unmounted it before ejecting, regardless of which method I use to eject the media. I take note of Keith's suggestion to eject from the command line in order to avoid the mount point remaining. The command 'eject /dev/sr0' does this well. Unfortunately, while the command opens the door, at this point I cannot mount anything because of the busy inodes problem, so I cannot test this to see if it really solves the problem of a mount point not being removed. It may also be easier to unmount from by right-clicking on the drive in "Places," as I recall Unmount is another option as well as Eject. I tried Dale's suggestion to use lsof and fuser, but no success so far. Lsof gives me pages of stuff I don't understand. I skimmed through the lsof man page looking for ways to limit the output to just things affecting /dev/sr0, but without success. Fuser, on the other hand, was much easier. I tried 'fuser -k /dev/sr0', which executed without error, but failed to do anything to solve the problem. I have a feeling that the busy inodes problem will disappear on rebooting. I resist that at the moment because I have programs running that I don't want to interrupt. More to the point, this may happen again, so having a way to fix the: [410083.768178] VFS: busy inodes on changed media or resized disk sr0 Without rebooting might be very useful. _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
