IBM 866011U - Netfinity 5500 - 8660, Hardware Maintenance Manual

Page: 128 / 402

F6: When a drive is removed from the
server (for example, for maintenance or
security), but is returned to a different bay,
the ServeRAID controller adjusts the
configuration to the drive's new location.
F7: A ServeRAID controller, with stored
configuration information that does not
match that of the configured drives present,
imports the configuration information from
the drives.
Ÿ
Read Ahead: The default setting is On.
Normally, the ServeRAID controller transfers
data from disk to its local cache in increments
equal to the stripe-unit size. This provides
excellent overall performance when workloads
are steady and sequential. However, if the
workload is random or the system I/O requests
are smaller than the stripe-unit size, reading
ahead to the end of the stripe might degrade
performance.
When read ahead is set to Off, the ServeRAID
controller transfers data from disk to its local
cache in increments equal to the system I/O
request size, without reading ahead to the end of
the stripe.
You can change the read-ahead setting without
destroying data in a logical drive.
4. Use the Up Arrow (
↑
), Down Arrow (
↓
), Right Arrow
(
→
), or Left Arrow (
←
) key to highlight the parameter
that you want to change; then, press Enter.
5. Press Esc to return to the Advanced Functions menu.
6. Back up the disk-array configuration information to
diskette. See “Backing up the disk-array
configuration” on page 126 for instructions. (The
backup procedure also saves the RAID parameters to
diskette.)
Changing the write policy:
When you configure a
logical drive, the ServeRAID controller sets the write policy
to
write-through (WT) mode, where the completion status
of a write command is sent after the data is written to the
hard disk drive. Under certain workloads, you can improve
performance by changing the write policy to
write-back
(WB) mode, where the completion status is sent after the
data is copied to cache memory, but before the data is
actually written to the storage device.
Although you might gain performance with write-back
mode, it creates a greater risk of losing data due to a
power failure. If a ServeRAID controller failure occurs
while data is in the cache the data is lost. This is because
the system gets a completion status message when the
data reaches cache memory, but
before data is actually
written to the hard disk drive.
120
Netfinity Server HMM