For anyone else who sees this thread: Do. Not. Enable. nobarrier. Barriers are important. If the device is able to use a non-volatile cache to speed up a sync, it should be able to do so anyway. -Sam On Tue, Jun 21, 2016 at 8:46 AM, Ning Yao <zay11022@xxxxxxxxx> wrote: > Hi, all > > As some RAID Controllers provides non-violate cache with battery to > protect data during power loss. For SATA disk, we want enable > nobarrier options to boost IO performance. But no barrier option > require us to use a real File on filesystem instead of a raw device, > which may cause more inode content updating and degrade performance. I > wonder, is that meaningful to add an option such like > enable_journal_nobarrier to tell whether open raw device with O_DSYNC > or not (default set to false), to guide the journal write without > barrier if we are definitely sure that our cache on SSD or RAID > Controller is non-violate? > > Regards > Ning Yao > -- > To unsubscribe from this list: send the line "unsubscribe ceph-devel" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
