I wasn't sure what location to post this so if there is a more appropriate location please move as seen fit. Thanks.


The closest I've been able to find around here is: http://www.linuxquestions.org/questi...xattr+returned
which is not quite what I currently have. Searching elsewhere online I haven't really been able to pin down my situation probably due to query shortcomings on my part.




I recently ran into a problem that I was unable to recover from without rebooting.

I started copying a large set of files (~1.4G worth of ~14-15MB files with a few <5KB files) into a directory. While this was copying (and I believe this was the cause of the hang) the system became non-responsive and I lost the ability to log in to any other shells (virtual-consoles, ttys, ...) to try to identify the problem and correct. After power-cycling and bringing everything up I noticed that the directory I created and was copying into now has no permissions on it and I am not able to remove it.

ls -l output for what should be a directory
and the directory 20080218 is blocked red, flashing directory name in grey (in case it matters)

in addition to that output of ls for this directory I see this entry in messages corresponding to it


I have since created another directory 20080218-0 and successfully copied the files to it so the data is not a concern for me.

Another note, this directory is in the root of the mounted filesystem.


I'm sure there is a way to fix (remove) this but I just have not been able to find how. I have some speculation as to what may need to be done to correct it (theory not know-how) but would like to refrain from posting that so as not to focus anyones attention to it but rather leave it open for what you may think.



Have any of you experienced this? Were you able to remove it? How did you fix (remove) it?


If you would like more information if I left anything out please don't hesitate to ask and I'll provide what I can.


Thanks,

If the partition that directory is on can be unmounted I'd run fsck over it. Otherwise, re-boot with a rescue CD and run fsck on it. If it's something corrupted in the file system, that should fix it. Otherwise can you post any error messages that are displayed during the check?

thanks. and boy is my face red...

i sat long enough to make sure that it started a forcefsck after passing the kernel option via grub from my power-cycle and then walked away; came back later to a login, figured it'd done its thing on all my filesystems and ready to go since didnt see any maintenance shell or error on the display.

sure enough though, i'd forgotten i set the fs_passno option in /etc/fstab to 0 for the filesystems on this disk and i can't for the life of me remember why. stopped/killed all the processes accessing the one with the problem in question, fsck'd it, fixing (deleting) the lost/orphaned/dereferenced inodes (the one referenced in messages), mounted it up and problem solved.

teach me to assume. and you can bet the fs_passno values have been set to be checked for them.

just needed a little push to question myself and what was actually done vs what i thought was actually done


Thanks again gilead!

No problem - glad you got it working

I've been caught in a similar way, firing off processes and going to get caffeine while they (supposedly) run...