Issue being that ideally, I'd like to take advantage of the logic built into Hadoop and Kafka to leverage JBOD for parallel I/O. My worry is that zfs drivers won't be as clever with application specific I/O optimizations. Z
Sent from my iPhone > On Apr 15, 2014, at 6:22 AM, Nux! <n...@li.nux.ro> wrote: > >> On 15.04.2014 14:03, Shanker Balan wrote: >> Comments inline. >>> On 07-Apr-2014, at 10:38 pm, Zack Payton <zpay...@gmail.com> wrote: >>> Damn, was hoping for JBOD. >>> Additionally, I don't have a raid controller. >>> Thanks anyhow. >> Zack, you can use Linux software RAID to create your RAID10 stripe >> and mount the volume under /var/lib/libvirt. >> http://www.tldp.org/HOWTO/Software-RAID-HOWTO.html > > Yes, good, also LVM over the disks would work, too. > > Lucian > > -- > Sent from the Delta quadrant using Borg technology! > > Nux! > www.nux.ro
