Re: [ClusterLabs] different start/stop order
On Fri, 2017-10-27 at 14:35 +0200, Stefan Krueger wrote: > Hello, > > is it possible to start a ressource in a different way then they > stopped? > eg: > primitive HA_IP-Serv1 IPaddr2 \ > params ip=172.16.101.70 cidr_netmask=16 \ > op monitor interval=20 timeout=30 on-fail=restart nic=bond0 \ > meta target-role=Started > primitive HA_IP-Serv2 IPaddr2 \ > params ip=172.16.101.74 cidr_netmask=16 \ > op monitor interval=10s nic=bond0 > primitive nc_storage ZFS \ > params pool=nc_storage importargs="-d /dev/disk/by- > partlabel/" > primitive nfs-serv1 systemd:nfs-server > primitive nfs-serv2 systemd:nfs-server > group compl_zfs-serv1 nc_storage nfs-serv1 HA_IP-Serv1 > > I want that pacemaker stop it in this order: > HA_IP-Serv1 nfs-serv1 nc_storage > > but start it in this order: > nfs-serv1 nc_storage HA_IP-Serv1 > > it that possible and how can I achieve that? > > best regards > Stefan Yes. You'd have to get rid of the group (which always starts in the listed order and stops in reverse), and instead use colocation and ordering constraints. You can make ordering constraints asymmetrical, so they only apply in the listed direction: http://clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html-single/Pacemake r_Explained/index.html#s-resource-ordering -- Ken Gaillot ___ Users mailing list: Users@clusterlabs.org http://lists.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
Re: [ClusterLabs] different start/stop order
Hello Ken, thanks for this information! > You can make ordering constraints asymmetrical, > so they only apply in the listed direction: > > http://clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html-single/Pacemaker_Explained/index.html#s-resource-ordering Ok but i see only how can i create a start order, but how can i create a different stop order? Best regards Stefan ___ Users mailing list: Users@clusterlabs.org http://lists.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
Re: [ClusterLabs] different start/stop order
On Fri, 2017-10-27 at 22:08 +0200, Stefan Krueger wrote: > Hello Ken, thanks for this information! > > > You can make ordering constraints asymmetrical, > > so they only apply in the listed direction: > > > > http://clusterlabs.org/doc/en-US/Pacemaker/1.1-pcs/html-single/Pace > > maker_Explained/index.html#s-resource-ordering > > Ok but i see only how can i create a start order, but how can i > create a different stop order? > > Best regards > Stefan symmetrical=false plus first-action and then-action -- Ken Gaillot ___ Users mailing list: Users@clusterlabs.org http://lists.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
Re: [ClusterLabs] different start/stop order
Hello Ken, thanks for you help, at the moment it looks like this (its not final): crm(live)# conf sho node 1084815873: zfs-serv1 node 1084815874: zfs-serv2 primitive HA_IP-Serv1 IPaddr2 \ params ip=172.16.101.70 cidr_netmask=16 \ op monitor interval=20 timeout=30 on-fail=restart nic=bond0 \ meta target-role=Started primitive HA_IP-Serv2 IPaddr2 \ params ip=172.16.101.74 cidr_netmask=16 \ op monitor interval=10s nic=bond0 primitive nc_storage ZFS \ params pool=nc_storage importargs="-d /dev/disk/by-partlabel/" primitive nfs-serv1 systemd:nfs-server primitive nfs-serv2 systemd:nfs-server group compl_zfs-serv2 nfs-serv2 HA_IP-Serv2 location cli-prefer-HA_IP-Serv1 HA_IP-Serv1 role=Started inf: zfs-serv1 location cli-prefer-HA_IP-Serv2 compl_zfs-serv2 role=Started inf: zfs-serv2 location cli-prefer-compl_zfs-serv1 nc_storage role=Started inf: zfs-serv1 location cli-prefer-nc_storage nc_storage role=Started inf: zfs-serv1 colocation compl_zfs-serv1 +inf: HA_IP-Serv1 nc_storage nfs-serv1 order start_serv1 nfs-serv1:start nc_storage:start HA_IP-Serv1:start symmetrical=false order stop_serv1 HA_IP-Serv1:stop nfs-serv1:stop nc_storage:stop symmetrical=false property cib-bootstrap-options: \ have-watchdog=false \ dc-version=1.1.16-94ff4df \ cluster-infrastructure=corosync \ cluster-name=debian \ no-quorum-policy=ignore \ default-resource-stickiness=100 \ stonith-enabled=false \ last-lrm-refresh=1507890181 is that ok? the manual failover looks good. best regards ___ Users mailing list: Users@clusterlabs.org http://lists.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
Re: [ClusterLabs] different start/stop order
On Wed, 2017-11-01 at 09:44 +0100, Stefan Krueger wrote: > Hello Ken, > > thanks for you help, at the moment it looks like this (its not > final): > crm(live)# conf sho > node 1084815873: zfs-serv1 > node 1084815874: zfs-serv2 > primitive HA_IP-Serv1 IPaddr2 \ > params ip=172.16.101.70 cidr_netmask=16 \ > op monitor interval=20 timeout=30 on-fail=restart nic=bond0 \ > meta target-role=Started > primitive HA_IP-Serv2 IPaddr2 \ > params ip=172.16.101.74 cidr_netmask=16 \ > op monitor interval=10s nic=bond0 > primitive nc_storage ZFS \ > params pool=nc_storage importargs="-d /dev/disk/by- > partlabel/" > primitive nfs-serv1 systemd:nfs-server > primitive nfs-serv2 systemd:nfs-server > group compl_zfs-serv2 nfs-serv2 HA_IP-Serv2 > location cli-prefer-HA_IP-Serv1 HA_IP-Serv1 role=Started inf: zfs- > serv1 > location cli-prefer-HA_IP-Serv2 compl_zfs-serv2 role=Started inf: > zfs-serv2 > location cli-prefer-compl_zfs-serv1 nc_storage role=Started inf: zfs- > serv1 > location cli-prefer-nc_storage nc_storage role=Started inf: zfs-serv1 FYI constraints that start with "cli-" were added automatically by command-line move/ban commands. They will stay in place (and effective) until you clear them. > colocation compl_zfs-serv1 +inf: HA_IP-Serv1 nc_storage nfs-serv1 > order start_serv1 nfs-serv1:start nc_storage:start HA_IP-Serv1:start > symmetrical=false > order stop_serv1 HA_IP-Serv1:stop nfs-serv1:stop nc_storage:stop > symmetrical=false I'm not expert with crm shell, so I could be wrong, but I thought that "colocation" and "order" lines could only list two resources. Otherwise this looks like it will do what you wanted. > property cib-bootstrap-options: \ > have-watchdog=false \ > dc-version=1.1.16-94ff4df \ > cluster-infrastructure=corosync \ > cluster-name=debian \ > no-quorum-policy=ignore \ > default-resource-stickiness=100 \ > stonith-enabled=false \ > last-lrm-refresh=1507890181 > > is that ok? the manual failover looks good. > > best regards -- Ken Gaillot ___ Users mailing list: Users@clusterlabs.org http://lists.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