Hardware Reference
In-Depth Information
FIGURE 11.4: Data and parity in N + M protect.
group can fail simultaneously and still provide 100% data availability. The M
erasure code members are computed from the N data members. Figure 11.4
shows the case for a regular 4 + 2 protection group (N = 4, M = 2, b = 1).
Because OneFS stripes files across nodes, this implies that files striped at
N+M can withstand M simultaneous node failures without loss of availability.
OneFS therefore provides resiliency across any type of failure, whether it be
to a drive, a node, or a component within a node (say, a card). Furthermore, a
node counts as a single failure, regardless of the number or type of components
that fail within it. Therefore, if five drives fail in a node, that still only counts
as a single failure for the purposes of N + M protection.
OneFS can uniquely provide a variable level of M, up to four, providing
for quadruple-failure protection. This goes far beyond the maximum level
of RAID commonly in use today, which is the double-failure protection of
RAID 6. Because the reliability of the storage increases geometrically with
this amount of redundancy, N + 4 protection can be hundreds or thousands
of times more reliable than traditional RAID 6. This added protection means
that large Serial AT Attachment (SATA) drives, such as 3 and 4 TB, can be
added with ease.
Smaller clusters can be protected with N + 1 protection, but this implies
that while a single drive or node could be recovered, two drives to two different
nodes could not. Drive failures are orders of magnitude more likely than node
failures. For clusters with large drives, it is desirable to provide protection
for multiple drive failures, though single-node recoverability is acceptable.
Fortunately, this can be provided very eciently on disk by using a value of
b > 1.
To provide for a situation where double-disk redundancy and single-node
redundancy is desired, one can set M = b = 2. Instead of creating protection
groups of N + M members, \double-length" protection groups are built up
