Hello again, > > > I'm unsure if the desync binary is also similarly broken when run in > > > an environment using ZFS storage. Performing the build in an identical > > > setup with ext4-backed storage works, so I'm not reporting this as an > > > RC bug. > > > > I wonder how I could reproduce this ? I guess I could setup a virtual > > machine > > with ZFS-backed storage. I've never played with ZFS, so not entirely sure > > what > > I am talking about here really. > > > > For completeness, can you confirm the version of ZFS you are using ? I > > assume it'll > > be the latest zfs-dkms (2.4.2-2) ? > > I managed to setup a Debian Forky VM with a separate 20G qcow disk and > format/mount it as ZFS. Wasn't too > much trouble. > > But the test suite (built the package with gbp buildpackage on the VM > directly) on both ext4 and ZFS > built and the tests ran fine. > > I wonder if my ZFS setup isn't as "complicated" as yours ? Perhaps the VM > adds some perfection into the mix ?
It turns out upstream has found the root-cause, described it very well in the GitHub issue and fixed the issue in https://github.com/folbricht/desync/pull/356 Once upstream releases a new version I will attempt to get it into Debian fairly quickly, or would you like it sooner and for me to backport the patch ? Thanks! Chris
