IBM Spectrum Scale (GPFS) for Linux on z SystemsThis IBM® RedpaperTM publication describes IBM Spectrum ScaleTM for Linux on z SystemsTM. This paper helps you install and configure IBM Spectrum Scale (formerly GPFSTM) in a disaster recovery configuration. Scenario testing is described for various events: Site failure, storage failure, node failure. Recovery procedures from each tested scenario are provided. This paper also provides an installation and configuration scenario for saving data stored in a Spectrum Scale file system by using IBM Spectrum ProtectTM integration features. Multi-node backup usage is described. |
Common terms and phrases
active undef running CHPID cio_ignore cluster configuration cluster file system cluster manager cluster nodes command DASD dasd_fs1 data deduplication dev/dasdc1 devices DEVTYPE ECKD fcp_fs1 forever preferred_lft forever free disk Fri Nov 20 GPFS PROXY HiperSockets IBM Redbooks IBM Spectrum Protect IBM Spectrum Scale inodes Linux guest lnxnode1 server lnxnode2 server lnxnode6 server node mmbackup mmchconfig mmcrnsd mmlsnsd multipath Node LinuxZ64 GPFS Node LinuxZ64 GPFS_PROXY node lnxnode1 node lnxnode5 server Node name node server node node7tie NSD disks nsd server access number of objects Pinging to check policy='service-time Processing disk ready up system Rediscovered nsd server renewal is overdue running active undef scenario SCSI server node lnxnode2 server node server Session shown in Example site_A_dasd_3 Spectrum Scale cluster Storage Manager server SUBCHANNEL SUSE Linux system desc Tcp/Ip Tcp/Ip Idlew Tcp/Ip Run Thu Dec Tivoli Storage Manager Total number TSMGPFS undef running active usage=dataAndMetadata valid_lft forever preferred_lft Yes Yes ready z/VM