[Debian-ha-maintainers] Bug#600606: pacemaker: Cluster dead after upgrade from madkiss to lenny-backports packages

[Michael Gebetsroither]

Package: pacemaker
Version: 1.0.9.1+hg15626-1~bpo50+1
Severity: normal

Hi,

After upgrading a test-cluster from madkiss packages to lenny-backports
pacemaker the whole cluster is dead.
corosync-cfgtool -s shows both rings are available, but crm_mon can not
connect to socket (/var/run/crm is empty and no other cluster daemon
beside corosync itself and ha_logd is running).

The whole thing is fixed temporarily with killall -9 corosync and
/etc/init.d/corosync start on all nodes.

real fix:
It seems the problem are strange startlevels of corosync after upgrade.
this fixed the problems here:
% update-rc.d -f corosync remove
 Removing any system startup links for /etc/init.d/corosync ...
   /etc/rc0.d/K07corosync
   /etc/rc1.d/K07corosync
   /etc/rc6.d/K07corosync
   /etc/rcS.d/S59corosync
% update-rc.d corosync defaults 
 Adding system startup for /etc/init.d/corosync ...
   /etc/rc0.d/K20corosync -> ../init.d/corosync
   /etc/rc1.d/K20corosync -> ../init.d/corosync
   /etc/rc6.d/K20corosync -> ../init.d/corosync
   /etc/rc2.d/S20corosync -> ../init.d/corosync
   /etc/rc3.d/S20corosync -> ../init.d/corosync
   /etc/rc4.d/S20corosync -> ../init.d/corosync
   /etc/rc5.d/S20corosync -> ../init.d/corosync

greets,
michael

-- System Information:
Debian Release: squeeze/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.35-trunk-amd64 (SMP w/2 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/bash