Hi, Have anyone configured 2 nodes cluster with corosync + pacemaker, where the nodes are in different subnet/network?
If the bindnetaddr differ, will the nodes join in the cluster? Is this feasible with corosync? Thanks, Raffi > -----Original Message----- > From: [email protected] [mailto:linux-ha- > [email protected]] On Behalf Of S, MOHAMED (MOHAMED)** CTR ** > Sent: Friday, September 21, 2012 9:08 PM > To: General Linux-HA mailing list > Subject: Re: [Linux-HA] corosync - nodes in different network > > Hi Dan, Emmanuel, > > Thanks for the quick responses. > > We have two variations. > 1- High Availability - Both the nodes will be in the same network in the > same lab. In this we have two physical interfaces, one through the > network, another one connected back to back. > > 2- Disaster Recovery - Both the nodes will be in different network located > in different places. In this case we have only one physical interface. > > Both used to work with heartbeat + pacemaker. > We are upgrading to corosync + pacemaker. > > First setup worked. Now, I am trying to configure the Disaster Recovery > setup with corosync + pacemaker. > > With Corosync, Is it not possible to have two nodes in different network > in a cluster? > > Thanks, > Raffi > > > > -----Original Message----- > > From: [email protected] [mailto:linux-ha- > > [email protected]] On Behalf Of Dan Frincu > > Sent: Friday, September 21, 2012 6:54 PM > > To: General Linux-HA mailing list > > Subject: Re: [Linux-HA] corosync - nodes in different network > > > > Hi, > > > > On Fri, Sep 21, 2012 at 4:17 PM, Emmanuel Saint-Joanis > > <[email protected]> wrote: > > > right this kind for example : > > > > More like this. > > > > interface { > > # The following values need to be set based on your > > environment > > ringnumber: 0 > > bindnetaddr: 172.16.17.18 > > mcastaddr: 226.94.1.1 > > mcastport: 5405 > > } > > interface { > > # The following values need to be set based on your > > environment > > ringnumber: 1 > > bindnetaddr: 192.168.169.170 > > mcastaddr: 226.94.1.2 > > mcastport: 5405 > > } > > > > There are a couple of issues with your setup (conceptually speaking). > > > > > > > > totem { > > > version: 2 > > > secauth: off > > > threads: 0 > > > rrp_mode: passive > > > interface { > > > ringnumber: 0 > > > member { > > > memberaddr: 10.0.0.11 > > > } > > > member { > > > memberaddr: 10.0.0.12 > > > } > > > bindnetaddr: 10.0.0.0 > > > mcastaddr: 226.94.1.1 > > > > mcastaddr does not apply to udpu transport. > > > > > mcastport: 5405 > > > ttl: 1 > > > } > > > interface { > > > ringnumber: 1 > > > member { > > > memberaddr: 192.168.0.11 > > > } > > > member { > > > memberaddr: 192.168.0.12 > > > } > > > bindnetaddr: 192.168.0.0 > > > mcastaddr: 226.94.1.2 > > > mcastport: 5407 > > > ttl: 1 > > > } > > > > And you're missing transport: udpu before the closing of the totem > stanza. > > > > > } > > > > > > > > > 2012/9/21 Dan Frincu <[email protected]> > > >> > > >> Hi, > > >> > > >> On Fri, Sep 21, 2012 at 2:51 PM, S, MOHAMED (MOHAMED)** CTR ** > > >> <[email protected]> wrote: > > >> > Hi, > > >> > If I set up the two nodes cluster in the same network (10.10.0.0, > > >> > 10.10.0.0), the nodes are joining in the cluster, I see the > > >> > "pcmk_peer_update" and "crm_update_peer" messages in > > >> > /var/log/cluster/corosync.log > > >> > > > >> > When I setup two nodes cluster with each node in different > > >> > network(10.10.0.0, 192.168.0.0), the nodes are not joining in the > > cluster. > > >> > Both the nodes has same /etc/corosync/authkey (confirmed through > > md5sum) > > >> > The nodes are not joining in the cluster; I do not see the > > >> > "pcmk_peer_update" messages in /var/log/cluster/corosync.log > > >> > I think I am not configuring the corosync.conf properly in this > > >> > scenario. > > >> > > > >> > The details of the two nodes are mentioned below. > > >> > Any help is really appreciated. > > >> > > >> First, both nodes for the same interface statement should be in the > > >> same network in order to work. > > >> Second, you need at least 2 physical interfaces for redundancy (and > > >> I'm not talking about bond here). You could set the 10.10.0.0/16 > > >> network on ringnumber 0 and 192.168.0.0/24 on ringnumber 1 (again, > > >> you'd need 2 physical interfaces). > > >> Third, ttl=1 for different subnets, how do you expect to route the > > >> packets? > > >> Fourth, my personal favourite, secauth=on and threads=0. Set > > >> threads=#number_of_cpus_on_the_system > > >> > > >> HTH, > > >> Dan > > >> > > >> > > > >> > Node A > > >> > IP: 10.10.0.38 > > >> > Netmask: 255.255.0.0 > > >> > corosync.conf > > >> > ================= > > >> > compatibility: whitetank > > >> > > > >> > totem { > > >> > version: 2 > > >> > secauth: on > > >> > threads: 0 > > >> > interface { > > >> > ringnumber: 0 > > >> > bindnetaddr: 10.10.0.0 > > >> > mcastaddr: 226.94.1.3 > > >> > mcastport: 3300 > > >> > ttl: 1 > > >> > } > > >> > } > > >> > > > >> > logging { > > >> > fileline: off > > >> > to_stderr: no > > >> > to_logfile: yes > > >> > to_syslog: yes > > >> > logfile: /var/log/cluster/corosync.log > > >> > debug: off > > >> > timestamp: on > > >> > logger_subsys { > > >> > subsys: AMF > > >> > debug: off > > >> > } > > >> > } > > >> > > > >> > amf { > > >> > mode: disabled > > >> > } > > >> > > > >> > service { > > >> > # Load the Pacemaker Cluster Resource Manager > > >> > name: pacemaker > > >> > ver: 0 > > >> > } > > >> > > > >> > Node B > > >> > IP: 192.168.0.199 > > >> > Netmask: 255.255.255.0 > > >> > corosync.conf > > >> > ================= > > >> > compatibility: whitetank > > >> > > > >> > totem { > > >> > version: 2 > > >> > secauth: on > > >> > threads: 0 > > >> > interface { > > >> > ringnumber: 0 > > >> > bindnetaddr: 192.168.0.0 > > >> > mcastaddr: 226.94.1.3 > > >> > mcastport: 3300 > > >> > ttl: 1 > > >> > } > > >> > } > > >> > > > >> > logging { > > >> > fileline: off > > >> > to_stderr: no > > >> > to_logfile: yes > > >> > to_syslog: yes > > >> > logfile: /var/log/cluster/corosync.log > > >> > debug: off > > >> > timestamp: on > > >> > logger_subsys { > > >> > subsys: AMF > > >> > debug: off > > >> > } > > >> > } > > >> > > > >> > amf { > > >> > mode: disabled > > >> > } > > >> > > > >> > service { > > >> > # Load the Pacemaker Cluster Resource Manager > > >> > name: pacemaker > > >> > ver: 0 > > >> > } > > >> > > > >> > Thanks, > > >> > Raffi > > >> > _______________________________________________ > > >> > Linux-HA mailing list > > >> > [email protected] > > >> > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > >> > See also: http://linux-ha.org/ReportingProblems > > >> > > >> > > >> > > >> -- > > >> Dan Frincu > > >> CCNA, RHCE > > >> _______________________________________________ > > >> Linux-HA mailing list > > >> [email protected] > > >> http://lists.linux-ha.org/mailman/listinfo/linux-ha > > >> See also: http://linux-ha.org/ReportingProblems > > > > > > > > > > > > > > -- > > Dan Frincu > > CCNA, RHCE > > _______________________________________________ > > Linux-HA mailing list > > [email protected] > > http://lists.linux-ha.org/mailman/listinfo/linux-ha > > See also: http://linux-ha.org/ReportingProblems > _______________________________________________ > Linux-HA mailing list > [email protected] > http://lists.linux-ha.org/mailman/listinfo/linux-ha > See also: http://linux-ha.org/ReportingProblems _______________________________________________ Linux-HA mailing list [email protected] http://lists.linux-ha.org/mailman/listinfo/linux-ha See also: http://linux-ha.org/ReportingProblems
