Anyone that follows this mailing list at all has probably noticed that I’m creating a 2-node HA iSCSI Target on RHEL 6 using Pacemaker/Corosync (and CMAN, I guess) and the available tgt scsi tools to use as shared file system for some Windows Server Failover Cluster nodes. After a great deal of trial and error, I’ve finally had success getting this to work, but I’m running into an issue with the file systems passing cluster validation on the Windows side. Initially, I was getting an error that the iSCSI target didn’t support SCSI-3 Persistent Reservations, which I was able to get around using the fence_scsi stonith module. I’ve found some extremely detailed conversations about the SCSI page 83h VPD descriptor error on the internet, but, frankly, I simply don’t follow them. Could someone ELI5 how to fix this **without scst, lio, lio-t**. Thanks!
# pcs status Cluster name: cluster Stack: cman Current DC: node1 (version 1.1.15-1.9a34920.git.el6-9a34920) - partition with quorum Last updated: Wed Sep 7 11:47:32 2016 Last change: Tue Sep 6 10:55:21 2016 by root via cibadmin on node1 2 nodes configured 8 resources configured Online: [ node1 node2 ] Full list of resources: Master/Slave Set: cluster-fs2o [cluster-fs1o] Masters: [ node1 ] Slaves: [ node2 ] cluster-vip (ocf::heartbeat:IPaddr2): Started node1 cluster-lvm (ocf::heartbeat:LVM): Started node1 cluster-tgt (ocf::heartbeat:iSCSITarget): Started node1 cluster-lun1 (ocf::heartbeat:iSCSILogicalUnit): Started node1 cluster-lun2 (ocf::heartbeat:iSCSILogicalUnit): Started node1 cluster-fence (stonith:fence_scsi): Started node2 PCSD Status: node1: Online node2: Online # cat /etc/cluster/cluster.conf <cluster config_version="9" name="cluster"> <fence_daemon/> <clusternodes> <clusternode name="node1" nodeid="1"> <fence> <method name="pcmk-method"> <device name="pcmk-redirect" port="node1"/> </method> </fence> </clusternode> <clusternode name="node2" nodeid="2"> <fence> <method name="pcmk-method"> <device name="pcmk-redirect" port="node2"/> </method> </fence> </clusternode> </clusternodes> <cman broadcast="no" expected_votes="1" transport="udpu" two_node="1"/> <fencedevices> <fencedevice agent="fence_pcmk" name="pcmk-redirect"/> </fencedevices> <rm> <failoverdomains/> <resources/> </rm> </cluster> # pcs config show Cluster Name: cluster Corosync Nodes: node1 node2 Pacemaker Nodes: node1 node2 Resources: Master: cluster-fs2o Meta Attrs: master-max=1 master-node-max=1 clone-max=2 clone-node-max=1 notify=true Resource: cluster-fs1o (class=ocf provider=linbit type=drbd) Attributes: drbd_resource=targetfs Operations: start interval=0s timeout=240 (cluster-fs1o-start-interval-0s) promote interval=0s timeout=90 (cluster-fs1o-promote-interval-0s) demote interval=0s timeout=90 (cluster-fs1o-demote-interval-0s) stop interval=0s timeout=100 (cluster-fs1o-stop-interval-0s) monitor interval=10s (cluster-fs1o-monitor-interval-10s) Resource: cluster-vip (class=ocf provider=heartbeat type=IPaddr2) Attributes: ip=10.30.96.100 cidr_netmask=32 nic=eth0 Operations: start interval=0s timeout=20s (cluster-vip-start-interval-0s) stop interval=0s timeout=20s (cluster-vip-stop-interval-0s) monitor interval=30s (cluster-vip-monitor-interval-30s) Resource: cluster-lvm (class=ocf provider=heartbeat type=LVM) Attributes: volgrpname=targetfs Operations: start interval=0s timeout=30 (cluster-lvm-start-interval-0s) stop interval=0s timeout=30 (cluster-lvm-stop-interval-0s) monitor interval=10s timeout=30 (cluster-lvm-monitor-interval-10s) Resource: cluster-tgt (class=ocf provider=heartbeat type=iSCSITarget) Attributes: iqn=iqn.2016-08.local.hsinauth.test:targetfs tid=1 incoming_username=iscsi incoming_password=@4TAt9-laObIrdeR Operations: start interval=0s timeout=10 (cluster-tgt-start-interval-0s) stop interval=0s timeout=10 (cluster-tgt-stop-interval-0s) monitor interval=10s timeout=20s (cluster-tgt-monitor-interval-10s) Resource: cluster-lun1 (class=ocf provider=heartbeat type=iSCSILogicalUnit) Attributes: target_iqn=iqn.2016-08.local.hsinauth.test:targetfs lun=1 path=/dev/targetfs/lun1 Operations: start interval=0s timeout=10 (cluster-lun1-start-interval-0s) stop interval=0s timeout=10 (cluster-lun1-stop-interval-0s) monitor interval=10 (cluster-lun1-monitor-interval-10) Resource: cluster-lun2 (class=ocf provider=heartbeat type=iSCSILogicalUnit) Attributes: target_iqn=iqn.2016-08.local.hsinauth.test:targetfs lun=2 path=/dev/targetfs/lun2 Operations: start interval=0s timeout=10 (cluster-lun2-start-interval-0s) stop interval=0s timeout=10 (cluster-lun2-stop-interval-0s) monitor interval=10 (cluster-lun2-monitor-interval-10) Stonith Devices: Resource: cluster-fence (class=stonith type=fence_scsi) Attributes: devices=/dev/targetfs/lun1,/dev/targetfs/lun2 Meta Attrs: provides=unfencing Operations: monitor interval=60s (cluster-fence-monitor-interval-60s) Fencing Levels: Location Constraints: Ordering Constraints: promote cluster-fs2o then start cluster-lvm (kind:Mandatory) (id:order-cluster-fs2o-cluster-lvm-mandatory) start cluster-vip then start cluster-lvm (kind:Mandatory) (id:order-cluster-vip-cluster-lvm-mandatory) start cluster-lvm then start cluster-tgt (kind:Mandatory) (id:order-cluster-lvm-cluster-tgt-mandatory) start cluster-tgt then start cluster-lun1 (kind:Mandatory) (id:order-cluster-tgt-cluster-lun1-mandatory) start cluster-tgt then start cluster-lun2 (kind:Mandatory) (id:order-cluster-tgt-cluster-lun2-mandatory) Colocation Constraints: cluster-vip with cluster-fs2o (score:INFINITY) (with-rsc-role:Master) (id:colocation-cluster-vip-cluster-fs2o-INFINITY) cluster-lvm with cluster-fs2o (score:INFINITY) (with-rsc-role:Master) (id:colocation-cluster-lvm-cluster-fs2o-INFINITY) cluster-tgt with cluster-fs2o (score:INFINITY) (with-rsc-role:Master) (id:colocation-cluster-tgt-cluster-fs2o-INFINITY) cluster-lun1 with cluster-fs2o (score:INFINITY) (with-rsc-role:Master) (id:colocation-cluster-lun1-cluster-fs2o-INFINITY) cluster-lun2 with cluster-fs2o (score:INFINITY) (with-rsc-role:Master) (id:colocation-cluster-lun2-cluster-fs2o-INFINITY) Resources Defaults: resource-stickiness: 100 Operations Defaults: No defaults set Cluster Properties: cluster-infrastructure: cman dc-version: 1.1.15-1.9a34920.git.el6-9a34920 default-resource-stickiness: 200 have-watchdog: false last-lrm-refresh: 1472233020 no-quorum-policy: ignore stonith-enabled: false -- [ jR ] there is no path to greatness; greatness is the path
_______________________________________________ Users mailing list: Users@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