On Tue, Jan 10, 2017 at 11:30:48PM +0100, Mattia Rizzolo wrote: > I still don't get this bug and why I can't reproduce it in any chroot > with eatmydata (and tbh it has very low priority in my task list, so > it's not like I ever looked at it since my last message), but anyway.
The bug should be reproducible with sbuild on a single CPU machine. I have described my building environment here: https://people.debian.org/~sanvila/my-building-environment.txt I changed my build setup since August, so to double check I did the following today: Built 23 times with eatmydata and overlayfs -> 23 build failures. Built 23 times with only overlayfs -> 23 successful builds. Built 23 times without any of the above -> 23 successful builds. (Each one of those 23 times was in a different machine). Compared with other things, I agree that this is low priority, but it seems that it's really eatmydata what makes it to fail. Thanks.
