> amd64, nothing special. > > Based on past experience, I suspect of some kind of weird interaction > between sbuild, eatmydata, systemd and/or the kernel from stretch, or maybe > the build process just does not fail gracefully on systems not having > a lot of memory (the bad build logs I have so far happened on machines > having 1GB RAM and 2GB swap).
Thanks for sharing your intuition on that. I'll have a look. F.
