User manual

rAid levelS
RAID arrays distribute data over multiple hard drives. RAID levels offer
different ways of managing the way data will be written to and read from
those hard drives. Some configurations are optimized for speed, others
for redundancy. To choose the best RAID level for your purposes, begin by
checking the standard practices for your field, and adjust to fit your specific
requirements. For detailed descriptions of common RAID levels supported
by the DataGuard Appliance, review the table on the next page.
Within each RAID level, you can configure stripe size and sector size.
Although most users do not need to change these settings, here is a brief
explanation of their role in balancing performance and redundancy The
RAID levels supported will depend on the number of hard drives in the
DataGuard Appliance. The table below shows the available RAID levels for
the number of hard drives installed.
Level Number of installed disks
1 2 3 4 5
RAID 0 X X X X X
RAID 1 X
RAID 10 X
RAID 1E X X X X
RAID 3 X X X
RAID 5 X X X
RAID 6 X X
Stripe Size
As files are accessed, they are broken up and written to the logical drive
in pieces called stripes. Because logical drives are made up of one or
more physical drives, a stripe must be divided again. This subdivided part
is called the stripe size. Put another way, a stripe size is the equal piece
written to each physical drive in a disk array; each and stripe size is part
of a stripe.
• When data is passed to the RAID controller, it is divided by increments
of the stripe size to create one or more blocks. These blocks are then
distributed among drives in the array, leaving different stripe-sized
pieces on different drives.
• Increasing the stripe size reduces the number of drives that a
given data file uses to hold all the blocks containing file data, while
decreasing the stripe size increases that number.
• In theory, by increasing the number of drives used will improve transfer
performance but diminish positioning performance. Increasing stripe
size requires fewer drives, and, in theory, reduces transfer performance
but improves positioning.
• Optimal stripe size depends on the user’s performance requirements
and what applications are running. Check the user documentation for
the software you will be using, and for the hard drives in the array.
SeCtor Size
A sector is the smallest storage unit used on a hard drive. A primary
consideration for determining what sector size to use is whether the logical
drive will be configured for NAS or as an iSCSI Target (LDType setting). For
iSCSI Target setups, the sector size should be optimized for the file system
with which it will be used. Consult the user documentation of the system
where it will be mounted and formatted.
reAd poliCy
The read policies determine if the controller reads sequential sectors of the
logical drive. Options are:
• ReadAhead: The controller reads sequential sectors of the logical drive.
Read-ahead policy can improve system performance if the data is
written to sequential sectors.
• ReadCache: The controller reads cache information to determine if the
data is available in the cache before retrieving the data from the disk.
This can provide faster read performance.
• NoCache: The controller retrieves data directly from the disk and not
from the cache.
write poliCy
Write policies determine if the controller sends a write-request completion
signal once the data is in the cache or until after it has been written to disk.
• WriteBack: The controller sends a write-request completion signal as
soon as the data is in the controller cache but has not yet been written
to disk. Write-back caching might improve performance but also entails
a data security risk since a system failure could prevent the data from
being written to disk and data might be lost. Other applications can
also experience problems when it is assumed that the requested data is
available on the disk.
• WriteThru: The controller sends a write-request completion signal only
after the data is written to the disk. Write-through caching provides
better data security than write-back caching but lower performance.
rebuilding A logiCAl drive
When a physical drive within a logical drive is identified as Dead and the
logical drive is identified as Critical, performing a rebuilding is necessary:
• For a fault tolerant RAID levels, (e.g. Raid 1, 5, 6, 10) if a spare drive is
available, the logical drive will begin rebuilding automatically.
• For a fault tolerant RAID levels, if no spare drive is available, you must
replace the failed hard drive. The logical drive will begin rebuilding
automatically, when you install the new hard drive.
• RAID 0 logical drives go offline after a physical drive failure. A RAID 0
logical drive cannot be rebuilt. All data on the logical drive is lost.
Notes: Rebuilding can take several hours depending on the size of your
physical drives. During a rebuild, you can access your folders on the
DataGuard Appliance.
34