On 12/07/11 03:36, Graham Rawolle wrote:
Hi all,
I am trying to run Corosync and Pacemaker onopenSUSE-11.4.
There are two machines, each with 2 network interfaces.
One interface on each is connected via a crossover cable on a private
network 192.168.100.0 and the other interface is connected via a
switch to our internal network 192.168.1.0.
Routing is set up so that 192.168.100.0/24 and multicast 224.0.0.0/4
is directed to eth1 (the private network).
*Question*- Does anyone have Corosync/Pacemaker running reliably on
openSUSE in a production environment? If so, what versions or
openSUSE, Corosunc and Pacemaker are you running?
Initially I was running Corosync 1.3.0-3.1-x86_64 and Pacemaker
1.1.5-3.2-x86_64. I managed to get Pacemaker started and partly
configured but ran in to 100% CPU issues, apparently due to a bug in
Corosync 1.3.0 which is fixed in version 1.3.2 or higher.
I have upgraded to Corosync 1.4.2-25.1 for openSUSE-11.4 from
build.opensuse.org, but now cannot get Pacemaker to start.
*Question*- Does Corosync 1.4.2 support configuring the Pacemaker
service with "ver:0"?
The openSUSE packaged versions of Pacemaker do not support configuring
the Pacemaker service with "ver:1" (pacemakerd and pacemaker startup
script are excluded).
Corosync seems to start and run fine. Both machines are seen and join
the membership.
I get the following message in /var/log/messages :
Dec 06 15:55:59 corosync [ SERV ] Service failed to load 'pacemaker'.
Dec 06 15:55:59 corosync [ SERV ] Service failed to load 'pacemaker'.
Note that it is repeated twice.
*Question - *How can I get more info on why Pacemaker failed to start?
I have the following versions installed:
openSUE 11.4-1.8-x86_64
Cluster-glue 1.07-9.1-x86_64
Corosync 1.4.2-25.1-x86_64
Pacemaker 1.1.5-3.2-x86_64
Openais 1.1.4-8.1-x86_64 (installed because Pacemaker requires it)
===================================
My /etc/corosync/corosync.conf file :
=============
# Please read the corosync.conf.5 manual page
compatibility: whitetank
totem {
version: 2
secauth: off
threads: 0
interface {
ringnumber: 0
bindnetaddr: 192.168.100.0
mcastaddr: 226.94.1.1
mcastport: 5405
ttl: 1
}
}
logging {
fileline: off
to_stderr: no
to_logfile: yes
to_syslog: yes
logfile: /var/log/cluster/corosync.log
debug: on
timestamp: on
logger_subsys {
subsys: AMF
debug: on
}
}
amf {
mode: disabled
}
=============
My /etc/corosync/service.d/pcmk file :
=============
service {
# Load the Pacemaker Resource Manager
name: pacemaker
ver: 0
}
=============
Regards,
Graham Rawolle
Daintree Systems
Mawson Lakes, South Australia
_______________________________________________
Pacemaker mailing list: Pacemaker@oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org
At one time I tried, but ran into quite a lot of problems. The SLES
versions seemed to be less burdened by these. My guess is that the real
development and testing (of course) happens there.
Personally, if money is involved, I think choosing those versions is
still the wise decision, as you waste consedirably less time.
I'm not with Novell by the way ;-)
B.
_______________________________________________
Pacemaker mailing list: Pacemaker@oss.clusterlabs.org
http://oss.clusterlabs.org/mailman/listinfo/pacemaker
Project Home: http://www.clusterlabs.org
Getting started: http://www.clusterlabs.org/doc/Cluster_from_Scratch.pdf
Bugs: http://bugs.clusterlabs.org