The mount -o remount,ro call does indeed not succeed. Essentially, the test ended up looking like this: fsck /ro (no errors) mount -o remount,rw /ro rsync stuff into /ro mount -o remount,ro /ro (this fails, /ro is still mounted as rw) fsck /ro (errors) During testing, I was able to determine that this is likely pointing us in the right direction. I tried a similar set of commands on a test system with aufs removed. In this system, the root is simply mounted read-only, and has some tmpfs directories mounted over certain areas to allow the system to write, I used a mount --bind to then ensure the rsync copies get put into the correct place. For simplicity, I used /ro as the bind location. fsck / (no errors) mount -o remount,rw / mount --bind / /ro rsync things into /ro fsck / (errors are reported) mount -o remount,ro / (succeeds, where the aufs version does not) fsck / (errors are no longer reported) So the difference would appear to be the remount command's failure.
On Thu, Aug 7, 2014 at 8:06 PM, <[1]sf...@users.sourceforge.net> wrote: Jacob Burkamper: > The fsck results are rather predictable, before the rsync there are not any > errors. After the rsync there are errors. Let me make sure. - your test script does  fsck /ro  mount -o remount,rw /ro  rsync something to /ro  mount -o remount,ro /ro  fsck /ro - first fsck doesn't report any error, but second fsck reports the  error. If so, did "remount,ro" really succeed? No error messages and you can confirm /ro is readonly via /proc/mounts? And your /sys/fs/aufs/si_*/xi_path shows "/rw/.aufs.xino"? J. R. Okajima -- Jacob Burkamper CIPAFilter Development Email: [2]jac...@cipafilter.com ---------------------- CIPAFilter Beta Program Email: [3]b...@cipafilter.com Web:     [4]http://www.cipafilter.com References 1. mailto:sf...@users.sourceforge.net 2. mailto:jac...@cipafilter.com 3. mailto:b...@cipafilter.com 4. http://www.cipafilter.com/
------------------------------------------------------------------------------ Want fast and easy access to all the code in your enterprise? Index and search up to 200,000 lines of code with a free copy of Black Duck Code Sight - the same software that powers the world's largest code search on Ohloh, the Black Duck Open Hub! Try it now. http://p.sf.net/sfu/bds