[Pacemaker] Can not use multicast, Any Ideas?

Dejan Muhamedagic dejanmm at fastmail.fm
Tue Jan 5 14:29:03 EST 2010


Hi,

On Tue, Jan 05, 2010 at 09:19:16PM +0200, Angie T. Muhammad wrote:
> Hello all
> 
> Thank you Dejan and Dr. Schwartzkopff
> But please bear with me because I'm still suffering a problem. Here is what
> I did:
> 
> #  wget -O /etc/yum.repos.d/clusterlabs.repo
> http://clusterlabs.org/rpm/epel-5/clusterlabs.repo
> # yum install pacemaker pacemaker-libs cluster-glue cluster-glue-libs
> resource-agents heartbeat
> =============================================================================================================================================================
>  Package                                    Arch
> Version                               Repository
> Size
> =============================================================================================================================================================
> Installing:
>  cluster-glue                               x86_64
> 1.0.1-1.el5                           clusterlabs
> 262 k
>  cluster-glue-libs                          x86_64
> 1.0.1-1.el5                           clusterlabs
> 130 k
>  heartbeat                                  x86_64
> 3.0.1-1.el5                           clusterlabs
> 193 k
>  pacemaker                                  x86_64
> 1.0.6-1.el5                           clusterlabs
> 689 k
>  pacemaker-libs                             x86_64
> 1.0.6-1.el5                           clusterlabs
> 310 k
>  resource-agents                            x86_64
> 1.0.1-1.el5                           clusterlabs
> 179 k
> Installing for dependencies:
>  corosync                                   x86_64
> 1.1.2-1.el5                           clusterlabs
> 163 k
>  corosynclib                                x86_64
> 1.1.2-1.el5                           clusterlabs
> 163 k
>  heartbeat-libs                             x86_64
> 3.0.1-1.el5                           clusterlabs
> 292 k
>  libesmtp                                   x86_64
> 1.0.4-5.el5                           epel
> 60 k
>  libibverbs                                 x86_64
> 1.1.2-4.el5                           base
> 44 k
>  librdmacm                                  x86_64
> 1.0.8-5.el5                           base
> 22 k
>  openhpi-libs                               x86_64
> 2.14.0-5.el5                          base
> 168 k
>  openib                                     noarch
> 1.4.1-3.el5                           base
> 20 k
> 
> Transaction Summary
> =============================================================================================================================================================
> Install     14 Package(s)
> Update       0 Package(s)
> Remove       0 Package(s)
> 
> Total download size: 2.6 M
> 
> # vim /etc/ha.d/ha.cf
> keepalive       2
> deadtime        30
> warntime        10
> initdead        120
> udpport         694
> ucast eth1      10.0.0.101
> auto_failback   on
> node            node1.mydomain.com
> node            node2.mydomain.com
> use_logd        yes
> 
> // and I changed the ucast directive properly for each node
> 
> # vim /etc/ha.d/authkeys
> # chmod 600 /etc/ha.d/authkeys
> # /etc/init.d/heartbeat start
> Starting High-Availability services:                       [  OK  ]
> // started properly on both nodes
> 
> # crm_mon -i5
> Attempting connection to the cluster....
> 
> # strace -o hb-again crm_mon -i5
> // the file is attached
> 
> // I didn't find perl on the system , so I installed it
> # yum install perl
> 
> // indeed, i believe the error is at around 92% of the strace output file
> when it attempts to:
> 
> connect(3, {sa_family=AF_FILE, path="/var/run/crm/cib_ro"...}, 110) = -1
> ENOENT (No such file or directory)
> close(3)                                = 0
> socket(PF_FILE, SOCK_STREAM, 0)         = 3
> fcntl(3, F_GETFL)                       = 0x2 (flags O_RDWR)
> fcntl(3, F_SETFL, O_RDWR|O_NONBLOCK)    = 0
> connect(3, {sa_family=AF_FILE, path="/var/run/crm/cib_callback"...}, 110) =
> -1 ENOENT (No such file or directory)

Looks like cib didn't start. The logs should say why. Perhaps
there are permission problems?

Thanks,

Dejan

> I can't understand why it can not run :( ..
> Version 1.0.5 of pace maker and openais 0.80.5 worked like a charm on the
> same nodes.
> Now I have to shift to heartbeat because of unicast directive. Please help!
> 
> Thank you in advance
> 
> 
> On Tue, Jan 5, 2010 at 2:17 PM, Michael Schwartzkopff <misch at multinet.de>wrote:
> 
> > Am Dienstag, 5. Januar 2010 13:00:44 schrieb Dejan Muhamedagic:
> > > Hi,
> > >
> > > On Tue, Jan 05, 2010 at 01:51:38PM +0200, Angie T. Muhammad wrote:
> > > > Hello all,
> > > > Hope you spent good time on holidays!
> > > >
> > > > Our data center does not support multicast and I have been googling
> > > > "unicast site:openais.org" but now results.
> > > > And changing our data center is not an option at the moment.
> > > >
> > > > I wonder does any beta version of openais support unicast?
> > >
> > > I think that the latest corosync (1.2.0) supports broadcast.
> > >
> > > > If not, do you have any link to pacemaker installation with heartbeat
> > > > stack?
> > >
> > > clusterlabs.org has some installation docs and there are also
> > > brand new docs at http://linux-ha.org/wiki/Documentation
> > >
> > > Thanks,
> > >
> > > Dejan
> > >
> > > > Indeed, I would be very grateful if you could suggest me any other
> > > > solution?
> >
> >
> > Perhaps you could use a tunnel (gre, ...) to route the multicast.
> >
> > --
> > Dr. Michael Schwartzkopff
> > MultiNET Services GmbH
> > Addresse: Bretonischer Ring 7; 85630 Grasbrunn; Germany
> > Tel: +49 - 89 - 45 69 11 0
> > Fax: +49 - 89 - 45 69 11 21
> > mob: +49 - 174 - 343 28 75
> >
> > mail: misch at multinet.de
> > web: www.multinet.de
> >
> > Sitz der Gesellschaft: 85630 Grasbrunn
> > Registergericht: Amtsgericht München HRB 114375
> > Geschäftsführer: Günter Jurgeneit, Hubert Martens
> >
> > ---
> >
> > PGP Fingerprint: F919 3919 FF12 ED5A 2801 DEA6 AA77 57A4 EDD8 979B
> > Skype: misch42
> >
> > _______________________________________________
> > Pacemaker mailing list
> > Pacemaker at oss.clusterlabs.org
> > http://oss.clusterlabs.org/mailman/listinfo/pacemaker
> >
> 
> 
> 
> -- 
> All the best,
> Angie


> _______________________________________________
> Pacemaker mailing list
> Pacemaker at oss.clusterlabs.org
> http://oss.clusterlabs.org/mailman/listinfo/pacemaker





More information about the Pacemaker mailing list