[Debian-ha-maintainers] Bug#804590: Bug#804590: redhat-cluster: build-depends on libcorosync-dev which is no longer built
Christoph Berg
myon at debian.org
Sat Jan 16 15:38:46 UTC 2016
Re: Ferenc Wagner 2016-01-16 <87k2n9ljk0.fsf at lant.ki.iif.hu>
> > libpengine4-dev (>= 1.1.8+git20121010)
> > libstonithd2-dev
> > libtransitioner2-dev
> > pacemaker-mgmt: pacemaker-dev (>= 1.1)
>
> Hmm, so that's why pacemaker hasn't migrated yet. I also spent some
> time with the tracker, and couldn't find anything but
> https://release.debian.org/migration/testing.pl?package=pacemaker with
> surprising statements like:
>
> pacemaker is not yet built on amd64: 1.1.13-2 vs 1.1.14~rc4-2 (missing 16 binaries)
There's pacemaker bits still on that version:
$ rmadison -S pacemaker
...
libpengine4 | 1.1.13-2 | unstable | amd64, arm64, armel, armhf, i386, mips, mips64el, mipsel, s390x
...
Decrufting unstable should take care of that, but we should probably
get more (all?) packages in shape first to make the ftp-team work
easier on that.
> corosync is only in unstable (no testing version)
No idea about that one. I'd hope it resolved itself when the other
problems are gone. (Testing propagation of bin-nmued packages isn't
really visible on DDPO et al, and as it doesn't involve updating to a
new source package in testing, there's surely tons of special code for
it in britney, of which we might be hitting a dozen corner cases.)
> On the other hand, I really wonder why crmsh does not depend on
> pacemaker (only recommends it). But we'll have to update it anyway.
I've just dirty-hacked its control file for the new pacemaker dev
packages, and it compiles. I'll upload it after some (superficial)
testing.
> By the way, Pacemaker 1.1.14 has been released, I'll fold it in
> shortly. And we'll also have to switch gfs2-utils and ocfs-tools to the
> new stack, so that redhat-cluster can finally be removed from the
> archive. If only dlm passed NEW already...
Would any of these be NEW as well? We can use the new dlm version to
upload more things to NEW. It's a bit inconvenient, but it should
speed up things.
Christoph
--
cb at df7cb.de | http://www.df7cb.de/
More information about the Debian-ha-maintainers
mailing list