At Thu, 25 May 2006 16:00:59 +0200, Tillmann Steinbrecher wrote: > On Tue, 9 May 2006 20:04:57 +0100, Alasdair G Kergon <agk@xxxxxxxxxx> wrote: > > On Mon, May 08, 2006 at 07:20:12PM +0200, Tillmann Steinbrecher wrote: > >> it's been many months that dm-crypt has been broken, and is known to > >> cause massive data corruption. > > I'm wondering if there is any news on the issue by > now. Unfortunately, I haven't found a way to reproduce it using loop > devices, but it is easy to reproduce using normal disks (although > the recent problems I had occured on a large RAID volume, I've also > had the problem in the past with a small 480 gig filesystem - so all > it takes to reproduce it is a few 160GB disks. Unfortunately, that's pretty much. We need a much much smaller test case. If it doesn't work with loop, try to put the whole thing in qemu and stress test it, maybe by just using bonnie++ multiple times. Maybe vmware. Maybe even UML? > The question we (pragmatic) users now have: Should we give up on > dm-crypt and use loop-aes, or can we expect a fix in the forseeable > future? I'm not sure what's with Christophe but I am putting any effort into this unless we have a test case that is in the magnitude of 10GB disk space. -- Fruhwirth Clemens - http://clemens.endorphin.org for robots: sp4mtrap@xxxxxxxxxxxxx --------------------------------------------------------------------- - http://www.saout.de/misc/dm-crypt/ To unsubscribe, e-mail: dm-crypt-unsubscribe@xxxxxxxx For additional commands, e-mail: dm-crypt-help@xxxxxxxx
