On Nov 21, 2018, at 22:32, Roger Sersted <[email protected]> wrote: > > > Lustre Server OS Version: CentOS 7.4 > Lustre Server version: 2.10.2 > Server Kernel version: 3.10.0-693.5.2 > Filesystem: ldiskfs > > I have an older disk array with a hardware failure. I suspect the > drives and data are still valid. > > What I plan to do is migrate the data off of a similar model array, move > the drives and migrate the data from the failed array. In order to do > this, I need to migrate just an OST to a different device. I can't use > lfs_migrate, because that works at the file level. I suspect the best > route to go is to create a new device and use "dd" to copy the contents > of the source device to the target device. > > Is there a better solution? > > Anything I should be aware of regarding dd and an OST? > > The arrays in question are enterprise class systems with caching > controllers configured for RAID6.
Hi Roger, I just noticed your email didn't get a response. The OST backup/restore process is discussed in the Lustre manual, https://build.whamcloud.com/job/lustre-manual/lastSuccessfulBuild/artifact/lustre_manual.html#dbdoclet.backup_device Using "dd" is fine, as long as the target device is as the source. Cheers, Andreas --- Andreas Dilger Principal Lustre Architect Whamcloud _______________________________________________ lustre-discuss mailing list [email protected] http://lists.lustre.org/listinfo.cgi/lustre-discuss-lustre.org
