Red Hat Bugzilla – Bug 877349
it should be possible to start corosync-notifyd without editing (nonexisting) /etc/sysconfig/corosync-notifyd
Last modified: 2013-11-20 23:31:51 EST
Description of problem: To actually be able to start corosync-notifyd without red FAIL one must do action similar to this: echo "OPTIONS=\"-d\" " > /etc/sysconfig/corosync-notifyd This should have been by default and user should be guided in comments in /etc/sysconfig/corosync-notifyd how to change this (dbus vs snmp). If sysconfig file cannot be provided, then the init script should be changed to include one of the usable options by default. Status message would probably be also welcome (either to stdout during service start and/or to syslog). Version-Release number of selected component (if applicable): corosync-1.4.1-11.el6.x86_64 How reproducible: always Steps to Reproduce: 1. clean install 2. /etc/rc.d/init.d/corosync-notifyd start 3. Actual results: Starting Corosync Notifier (corosync-notifyd): [FAILED] Syslog: Nov 16 04:11:36 marathon-03c1-node01 corosync-notifyd[21286]: no event type enabled, see corosync-notifyd -h, exiting. Expected results: service starts with reasonable defaults Additional info:
(In reply to comment #0) > Description of problem: > To actually be able to start corosync-notifyd without red FAIL one must do > action similar to this: > echo "OPTIONS=\"-d\" " > /etc/sysconfig/corosync-notifyd > > This should have been by default and user should be guided in comments in > /etc/sysconfig/corosync-notifyd how to change this (dbus vs snmp). > Agree > If sysconfig file cannot be provided, then the init script should be changed > to include one of the usable options by default. > > Status message would probably be also welcome (either to stdout during > service start and/or to syslog). > ???? As you noted later, there is "no event type enabled, see corosync-notifyd -h, exiting." message in syslog. What you mean by this requirement?
> > If sysconfig file cannot be provided, then the init script should be changed > > to include one of the usable options by default. > > > > Status message would probably be also welcome (either to stdout during > > service start and/or to syslog). > > > > ???? As you noted later, there is "no event type enabled, see > corosync-notifyd -h, exiting." message in syslog. What you mean by this > requirement? Something like "Using dbus for notifications. Use /etc/sysconfig/corosync-notifyd for override."?
Created attachment 736371 [details] Proposed patch - part 1 - Improve corosync-notifyd example Example now contains default option -d, so corosync-notifyd init script don't fall. Also description is improved a little bit.
Created attachment 736381 [details] Proposed patch - part 2 - Install sysconfig/corosync-notifyd in specfile
(In reply to comment #3) > > > If sysconfig file cannot be provided, then the init script should be changed > > > to include one of the usable options by default. > > > > > > Status message would probably be also welcome (either to stdout during > > > service start and/or to syslog). > > > > > > > ???? As you noted later, there is "no event type enabled, see > > corosync-notifyd -h, exiting." message in syslog. What you mean by this > > requirement? > > Something like "Using dbus for notifications. Use > /etc/sysconfig/corosync-notifyd for override."? I don't think this is really achievable in consistent way. First of all, from upstream point of view, corosync-notifyd init script is just init script. It's possible to run corosync-notifyd without that. And second problem is, what about debian? They have /etc/defaults (not /etc/sysconfig) so it mean to hack ether init script to display these message what means to parse OPTIONS what is just overkill (and extra useless work effort) or corosync-notifyd.c source code where we are back to /etc/defaults vs. /etc/sysconfig problem. Also I don't think such message is extremely useful. With patch 1, -d is now default so corosync-notifyd is started and works. Other then that is really about common sense (like almost all of daemons in RHEL has config in /etc/sysconfig/daemon_name) or documentation.
# service corosync-notifyd start Starting Corosync Notifier (corosync-notifyd): [ OK ] # Config file present in /etc/sysconfig/corosync-notifyd Verified with corosync-1.4.1-17.el6.x86_64. Thank you.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. http://rhn.redhat.com/errata/RHBA-2013-1531.html