Table of Contents
Advertisement
specify an Ultra Low value. Like volume initialization, after a volume rebuilds, it does not require a system
reboot.
High Reliability
Hard Drive Failure Prediction
•
In an effort to help users avoid data loss, disk manufacturers are now incorporating logic into their drives
that acts as an "early warning system" for pending drive problems. This system is called SMART. The disk
integrated controller works with multiple sensors to monitor various aspects of the drive's
performance, determines from this information if the drive is behaving normally or not, and makes available
status information to SAS RAID controller firmware that probes the drive and look at it.
The SMART can often predict a problem before failure occurs. The controllers will recognize a SMART error
code and notify the administrator of an impending hard drive failure.
Auto Reassign Sector
•
Under normal operation, even initially defect-free drive media can develop defects. This is a common
phenomenon. The bit density and rotational speed of disks is increasing every year, and so are the potential
of problems. Usually a drive can internally remap bad sectors without external help using cyclic redundancy
check (CRC) checksums stored at the end of each sector.
The RAID controller drives perform automatic defect reassignment for both read and write errors. Writes
are always completed - if a location to be written is found to be defective, the drive will automatically
relocate that write command to a new location and map out the defective location. If there is a recoverable
read error, the correct data will be transferred to the host and that location will be tested by the drive to be
certain the location is not defective. If it is found to have a defect, data will be automatically relocated, and
the defective location is mapped out to prevent future write attempts.
In the event of an unrecoverable read error, the error will be reported to the host and the location will be
flagged as being potentially defective. A subsequent write to that location will initiate a sector test and
relocation should that location prove to have a defect. Auto Reassign Sector does not affect disk subsystem
performance because it runs as a background task. Auto Reassign Sector discontinues when the operating
system makes a request.
Consistency Check
•
A consistency check is a process that verifies the integrity of redundant data. To verify RAID 3, 5, 6, 30, 50 or
60 redundancy, a consistency check reads all associated data blocks, computes parity, reads parity, and
verifies that the computed parity matches the read parity.
Consistency checks are very important because they detect and correct parity errors or bad disk blocks in
the drive. A consistency check forces every block on a volume to be read, and any bad blocks are marked;
those blocks are not used again. This is critical and important because a bad disk block can prevent a disk
rebuild from completing. We strongly recommend that you run consistency checks on a regular ba
sis—at least once per week. Note that consistency checks degrade performance, so you should run them
when the system load can tolerate it.
120
Advertisement
Table of Contents
