On Friday 24 April 2009 21:45:55 rpmfusion-buildsys(a)lists.rpmfusion.org wrote: > Job failed on arch x86_64 > > > Build logs may be found at http://buildsys.rpmfusion.org/logs/fedora-development-rpmfusion_free/3629... > > > ------------------------------------------------- [...] > {standard input}: Assembler messages: > {standard input}:51908: Fatal error: can't write moc_scanwizard.o: No space left on device > {standard input}:51908: Fatal error: can't close moc_scanwizard.o: No space left on device > make[2]: Leaving directory `/builddir/build/BUILD/mythtv-0.22/mythtv-0.22/libs/libmythtv' > make[2]: *** [moc_scanwizard.o] Error 1 > make[1]: Leaving directory `/builddir/build/BUILD/mythtv-0.22/mythtv-0.22/libs' > make[1]: *** [sub-libmythtv-make_default] Error 2 > make: *** [sub-libs-make_default-ordered] Error 2 > error: Bad exit status from /var/tmp/rpm-tmp.UJHjrN (%build) Seems the lisas.de builder is out of disk space... http://buildsys.rpmfusion.org/build-status/job.psp?uid=3629

> To we want mock to use caching? I'd say we want a root_cache, as it seems to speed things up a bit. I guess we could/should disable the yum_cache, as all builders have nearby mirrors. Or we enable it and stop maintaining those nearby mirrors. Or we chose depending on builder. I'm not sure if ccache is worth the trouble. I tend to say "no". Does anyone know if Fedora handles those different caching strategies on its builders? > Because then I have to increase the disk space (which is no problem). FYI, /var/cache/mock/ grew up to 13G on j-rod's builder over the past few days (which is not building EL). But I guess it'll grow even more over time.