Red Hat Bugzilla – Bug 227278
reiserfstune not found during install
Last modified: 2007-11-30 17:11:56 EST
While installing onto a system using a pre-existing reiserfs /home, anaconda
traces back because it can't find reiserfstune.
Traceback follows - full dump is attached.
Traceback (most recent call first):
File "/usr/lib/anaconda/iutil.py", line 51, in execWithRedirect
raise RuntimeError, "Error running " + command + ": " + msg
File "/usr/lib/anaconda/fsset.py", line 434, in labelDevice
stderr = "/dev/tty5", searchPath = 1)
File "/usr/lib/anaconda/fsset.py", line 1571, in labelEntry
File "/usr/lib/anaconda/fsset.py", line 1688, in makeFilesystems
File "/usr/lib/anaconda/packages.py", line 144, in turnOnFilesystems
File "/usr/lib/anaconda/dispatch.py", line 201, in moveStep
rc = stepFunc(self.anaconda)
File "/usr/lib/anaconda/dispatch.py", line 124, in gotoNext
File "/usr/lib/anaconda/gui.py", line 973, in nextClicked
File "/usr/lib/anaconda/iw/progress_gui.py", line 247, in renderCallback
File "/usr/lib/anaconda/gui.py", line 1000, in handleRenderCallback
RuntimeError: Error running reiserfstune: No such file or directory
Created attachment 147309 [details]
full anaconda traceback
We should make it so that we handle the tools for various filesystems not being
reiserfstune should be there. scripts/upd-instroot has reiserfs-utils in the
pile of packages to use to build the stage2 image.
Still, I agree with Jeremy, execWithRedirect (among others) could gracefully
handle a tool missing.
What are the exact install parameters here? reiserfs on a /home partition, no
existing other installs or partitions?
I believe there was an existing install on the disk that was being replaced with
F7t1, where /home was reiserfs and the other partitions were (probably) ext3 -
IIRC the user was attempting to install (not upgrade) but also wanted to
I'm unable to reproduce this today. Perhaps there was some transient tree
weirdness that caused reiserfstune to not be in the image?
Possibly. Does reiserfstune live in initrd or stage2? We should compare the
current rawhide one to the F7t1 one and see if the location (or presence) of the
file has changed.
Looks like it was left out of the minstg2.img on Test 1, but it's in the images
now. Closing out.
Just to clarify, the reiserfs tools didn't make the manifest for Test1 (and
possibly test2). Twas a pungi bug.