guzu
asked on
cpio:lstat failed AND cpio: mkdir failed
i use ximian desktop on my red hat linux 7.2 machine. ximian corporation released an new version of their software downloader, red-carpet. when i started red-carpet, he says that i must upgrade my (old) red-carpet with the new one. he keeps saying this and not did not let me do anything else. the problem is that upgrading (rpm -Uvh) crashes with the error "unpacking of archive failed: cpio: lstat failed - input/output error". i thought that maybe rpm was too old for that package. i have tried to upgrade rpm-4.0.3 to rpm-4.0.4. : "error: unpacking of archive failed: cpio: mkdir failed - input/output error". this is happening while i'm logged in as root. what's wrong?
guzu
guzu
The 7.2 version of rpm isn't the problem here. It works fine with Ximian. Do you have enough free space on the system for the unpacking of the rpm? Do a 'df' and see what your filesystems look like.
ASKER
hello jlevie, i was waiting for you. df -h says i have 297 megs free. also i have 1gig of free swap memory. unpacking other packages was ok.
ASKER
red-carpet was downloaded several times and the error was the same: "cpio: lstata failed"....
Hmm, I've not encounted that error with Ximian. Are you using Ximian's tool to do the upgrade of RedCarpet? Normally it notices that it's version is old and asks to be allowed to do the update. I don't know what would happen if you tried to manually do the upgrade via an rpm.
ASKER
well, that's how the all thing's started. i've tried to upgrade with ximian's tool. indeed, it asks to do the update. as you know, there's no other posibility so i pressed the update button. he says that everything is ok and it will restart i have to agree with that then he started over again and again. the i tried to do it manually, and i've found that errors. i've checked for the installed version of red-carpet and it was the old one... . i thought that it was rpm's fault (old rpm utility with new rpm package..) and the rest you know. could it be the hdd?
ASKER
well, that's how the all thing's started. i've tried to upgrade with ximian's tool. indeed, it asks to do the update. as you know, there's no other posibility so i pressed the update button. he says that everything is ok and it will restart i have to agree with that then he started over again and again. the i tried to do it manually, and i've found that errors. i've checked for the installed version of red-carpet and it was the old one... . i thought that it was rpm's fault (old rpm utility with new rpm package..) and the rest you know. could it be the hdd?
ASKER
sorry, i've reposted data during refresh...
Besides Ximian, what other updates have been applied to the system and have all that are applied come from RedHat?
ASKER
all updates (maybe except libstdc3, i can't recall now) are from ximian or redhat
ASKER
bad news jlevie, i've done e2fsck -nf on my root partition and there are A LOT of errors... . do you think that e2fsck-ing from slackware would fix the errors? or i will loose information? (i have some bad memories about running e2fsck on my ext3 filesystem...
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
ASKER
well, last time i've tried this, i've lost everything so, i'll try to boot a slack and fsck without having the partition mounted.... thanks. i'll keep you up to date.
As far as I know it's safe to force the fsck at boot on a 7.x system. What you can't safely do is to manually fsck the root file system with it mounted and in use. The OS knows how to fsck the root at boot time, which is why the forcefsck technique exists.
Now, if you've got a flaky disk or controller all bets are off. The only times I've seen corruption of a file system involved bad disks/controllers or wierd software that ran with root privs. Under normal operation (oderly startup/shutdown of the OS) it's very unlikely for RH 7.2 to scramble anything on the system.
Now, if you've got a flaky disk or controller all bets are off. The only times I've seen corruption of a file system involved bad disks/controllers or wierd software that ran with root privs. Under normal operation (oderly startup/shutdown of the OS) it's very unlikely for RH 7.2 to scramble anything on the system.
ASKER
you are the expert onceagain. since i have strong reason to believe my hdd/mb don't match (for whatever reason), i've done the check from another partition/distribution. since i still can write you this, it seems to me that it works. also , the packages could now be installed on my system. probably, first step will be to replace the hdd. ok. now the points will be yours. thanks again.
I'm glad we go to the bottom of this. I'll admit to being a bit mystified as to the rpm failures until you mentioned the disk problem. At that point the cause of the failure became quite clear. I still don't know why the corruption occured, but a disk/mb mismatch could be the root cause.
ASKER