Hello, In our three-node GFS cluster, one node has a different GFS cookie from the other nodes for the same shared iSCSI LUN. [root:node11]# gfs_tool list 4172664832 sdb cluster1:iscsi.2 [root:node12]# gfs_tool list 4172664832 sdb cluster1:iscsi.1 [root:node13]# gfs_tool list 4171632640 sdb cluster1:iscsi.0 Is this a possible (correct) state? R/W is doable and gfs_fsck is successful on all nodes. But "gfs_tool freeze" does not accept a mountpoint as an argument (cookie is OK). Does this bug still remains in FC4's GFS packages? I'm using FC4-i386 2.6.11-1.1369_FC4smp and the following RPMs: (iSCSI initiator is SFnet linux-iscsi 4.0.2.1.test) GFS-6.1.0-3 GFS-kernel-smp-2.6.11.8-20050601.152643.FC4.2 GFS-kernheaders-2.6.11.8-20050601.152643.FC4.2 cman-1.0.0-1 cman-kernel-smp-2.6.11.5-20050601.152643.FC4.2 cman-kernheaders-2.6.11.5-20050601.152643.FC4.2 dlm-1.0.0-3 dlm-devel-1.0.0-3 dlm-kernel-smp-2.6.11.5-20050601.152643.FC4.2 dlm-kernheaders-2.6.11.5-20050601.152643.FC4.2 ... Things have not changed by re-gfs_mkfs or OS re-installation. Thanks, -- Kenji -- Linux-cluster@xxxxxxxxxx http://www.redhat.com/mailman/listinfo/linux-cluster
