[ClusterLabs] RES: Pacemaker issue lsb service

[Carlos Xavier]

Hi.

I dont know IF Ubuntu is system compliant, on OpenSuse Leap 42.1 I got it working.
First you have to create the configuration file, I got two files, since I listen to UDP and TCP to be Mikrotik compatible.

/etc/openvpn/openvpnTCP.conf
/etc/openvpn/openvpnUDP.conf

On systemd environment each instance of Openvpn is enabled using the configuration file name, so on Pacemaker we have this:

 primitive openvpnTCP systemd:openvpn at openvpnTCP \
        op start timeout=60 interval=0 \
        op stop timeout=60 interval=0 \
        op monitor interval=60 timeout=60
primitive openvpnUDP systemd:openvpn at openvpnUDP \
        op start timeout=60 interval=0 \
        op stop timeout=60 interval=0 \
        op monitor interval=60 timeout=60

group openvpn-group openvpnUDP openvpnTCP


Also, if you do not start any connection from your Openvpn i.e. you just receive connections from clients you can start Openvpn on both servers, letting the tunnels ready for connection.

clone openvpn-clone openvpn-group \
        meta target-role=Started globally-unique=false

Regards,
Carlos Xavier.



> -----Mensagem original-----
> De: Kristoffer Grönlund [mailto:kgronlund at suse.com]
> Enviada em: segunda-feira, 7 de março de 2016 18:37
> Para: Thorsten Stremetzne; users at clusterlabs.org
> Assunto: Re: [ClusterLabs] Pacemaker issue lsb service
> 
> Thorsten Stremetzne <clusterlabs at webants.com> writes:
> 
> > Hello all,
> >
> >
> > I have built an HA setup for a OpenVPN server.
> > In my setup there are two hosts, running Ubuntu Linux, pacemaker &
> > chorosync. Also both hosts have a virtual IP which migrates to the
> > host that is active, when the other fails. This works well, but I also
> > configured a primitive for the openvpn-server init scrip, via
> >
> >
> > crm configure primitive failover-openvpnas lsb::openvpnas op monitor
> > interval=15s
> >
> 
> Hi,
> 
> Unfortunately most LSB init scripts are not cluster-compatible by default, they often do not implement
> monitor actions correctly and may report incorrect status when the resource is not running.
> 
> I would recommend using an OCF resource agent if possible, or worst case wrapping the LSB init script
> in a custom OCF resource agent which handles the corner cases. Another option if you are running a
> system with systemd is to use a systemd service. I have heard reports that there are some issues with
> using systemd services directly as well, but the ones I have tried have worked out of the box.
> 
> Cheers,
> Kristoffer
> 
> >
> > The service will be added, but it will always fail, due to the syslog, the init script will be
> called in a wrong way.
> > I'm in troubles debugging how pacemaker will try to start/stop the service on the hosts.
> >
> >
> > Can someone please assist me with some ideas and suggestions?
> >
> >
> > Thanks very much
> >
> >
> > Thorsten
> >
> >
> > Diese E-Mail kann vertrauliche und/oder rechtlich geschützte Informationen enthalten. Wenn Sie nicht
> der beabsichtigte Empfänger sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte
> sofort den Absender telefonisch oder per E-Mail und löschen Sie diese E-Mail aus Ihrem System. Das
> unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet.
> > _______________________________________________
> > Users mailing list: Users at clusterlabs.org
> > http://clusterlabs.org/mailman/listinfo/users
> >
> > Project Home: http://www.clusterlabs.org Getting started:
> > http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> > Bugs: http://bugs.clusterlabs.org
> 
> --
> // Kristoffer Grönlund
> // kgronlund at suse.com
> 
> _______________________________________________
> Users mailing list: Users at clusterlabs.org http://clusterlabs.org/mailman/listinfo/users
> 
> Project Home: http://www.clusterlabs.org Getting started:
> http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
> Bugs: http://bugs.clusterlabs.org