On Tue, 22 Sep 2009, Richard W.M. Jones wrote:
Previously (yum 3.2.24-4) yum ignored the case where some package fails to unpack, as in here: Installing : filesystem-2.4.30-2.fc12.x86_64 3/132 Error unpacking rpm package filesystem-2.4.30-2.fc12.x86_64 error: unpacking of archive failed on file /proc: cpio: utime [but yum continues anyway] However with the latest yum 3.2.24-5 it fails with this obscure stack trace, which we assume is related to the above error: Traceback (most recent call last): File "/usr/lib/python2.6/site-packages/yum/rpmtrans.py", line 377, in callback self._instCloseFile( bytes, total, h ) File "/usr/lib/python2.6/site-packages/yum/rpmtrans.py", line 452, in _instCloseFile state = self.base.history.txmbr2state(txmbr) File "/usr/lib/python2.6/site-packages/yum/history.py", line 280, in txmbr2state state = _stcode2sttxt[txmbr.output_state] KeyError: 100 error: python callback <bound method RPMTransaction.callback of <yum.rpmtrans.RPMTransaction instance at 0x3ae68c0>> failed, aborting! We'd like it to continue running in this case (or at least, for that to be made configurable) because the particular error here is inconsequential for us. This is the apparent patch between the two versions: http://cvs.fedoraproject.org/viewvc/devel/yum/yum-HEAD.patch?r1=1.32&r2=1.33&view=patch Lots of changes to do with the new history feature, and some changes to error handling, but I can't see anything obvious that would cause the above stack trace ...
Let's be clear - the traceback should be handled - however you had the cpio archive inside the rpm fail to unpack. That's not an error you should ignore. It means the bits did NOT get onto the system.
I agree we shouldn't traceback - and we'll get that cleaned up pronto - but I don't think we should be pleased to have an unpacking failure.
-sv _______________________________________________ Yum-devel mailing list Yum-devel@lists.baseurl.org http://lists.baseurl.org/mailman/listinfo/yum-devel
