As reported on Bugzilla, yum will fail silently if it is run from a rootsquashed nfs mount, since rpm will not have any read or write access. Due to a previous patch, yum will already correctly report that the operations fail, but it will not give any explanation as to why. With this patch, most operations should simply complete successfully, but note that changing the working directory will break things such as specifying rpms to install via relative paths.
Arguably, it might be better to instead just report to the user that rpm has run into errors since it doesn't have the necessary permissions, but ifaict there's not really any way to do that. When ts.run() is called both in testing the transaction and actually running the transaction it returns an empty list, which is supposed to indicate that it has experienced non-fatal errors. Since ifaict there's no way to tell what the errors are, in this case we can't report them to the user. Furthermore, the output of rpm is sent to a temporary file to be logged. It seems that the output can be displayed to the user via certain callbacks, but in this case rpm doesn't call any callbacks at all. So it seems that the temporary file is just deleted. _______________________________________________ Yum-devel mailing list [email protected] http://lists.baseurl.org/mailman/listinfo/yum-devel
