On 2013-05-15, at 5:50, Eitan Rosenfeld <eitan27@xxxxxxxxx> wrote: > I am working on improving the resiliency of a storage system using ext3. > > I will maintain two ext3 partitions of identical size on two different machines. Why not use e.g. DRBD or MD RAID1 with iSCSI targets on the remote nodes? > If each partition stores the same files, what are the sources of > bit-level inconsistencies between the two partitions? Is there any > nondeterminism? > > For example, between two otherwise identical partitions, the HTREE > hash seeds are different in the superblock. How might this impact the > downstream bits in the partitions? The hashing order will be completely different, so this will produce a completely different directory order. In ext4, the checksums also depend on the UUID, so these would also be different for each filesystem. Cheers, Andreas > You can assume that all files will be saved in the same order, and all > timestamps and stat output are identical between files. > > Thank you for your time, > Eitan > -- > To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" 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 linux-fsdevel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
