Configuring Two Serveraid Controllers In A Failover Environment (Windows Nt 4.0 Only); Hardware Requirements - IBM Netfinity ServeRAID-3H Installation And User Manual

Drive State
Blocked
Critical Migrating
Critical System
Critical
Migrating
Offline
Okay
System
Configuring two ServeRAID controllers in a failover environment
(Windows NT 4.0 only)
You can now configure two ServeRAID-3 controllers in a failover environment when using Windows NT
4.0. The latest version of the ServeRAID-3 device drivers includes fault tolerant technology. With fault
tolerance, you can pair two controllers and connect them to the same enclosure in order to provide access
to the disk, even after a controller fails.

Hardware requirements

Important
To perform fault tolerance, you must use only matched pairs of ServeRAID controllers. The controllers
must be identical, including the level of BIOS and firmware on the controllers.
You can use the following pairs for fault tolerance:
Two ServeRAID-3H
Meaning
During a rebuild operation, the ServeRAID controller sets the state of any RAID level-0
logical drives associated with a failed array to the Blocked state; then, it reconstructs the
data that was stored in RAID level-1, RAID level-1E, Raid level-5 and RAID level-5E
logical drives.
After the Rebuild operation completes, you can unblock the RAID level-0 logical drives,
and access them once again. However, the logical drive might contain damaged data.
You must either re-create, reinstall, or restore the data from the most recent backup disk
or tape to the RAID level-0 logical drive.
A logical drive in the critical state that is undergoing a logical drive migration (LDM).
The ServeRAID controller uses this reserved state during a logical drive migration (LDM)
when the logical drive is in the critical state.
A RAID level-1, RAID level-1E, RAID level-5, or RAID level-5E logical drive that contains
a defunct physical drive is in the critical state. A critical logical drive is accessible,
despite a physical drive failure.
Attention:
If the state of the logical drive is critical, refer to "Rebuilding a defunct drive" on
page 105.
The logical drive is undergoing a logical drive migration; that is, a change in RAID levels,
a change in logical drive size, an increase in free space, or a RAID level-5E compression
or decompression.
The logical drive is offline and not accessible. This state occurs when one of the
following is true.
One or more physical drives in a RAID level-0 logical drive is defunct.
Two or more physical drives in a RAID level-1, RAID level-1E, or level-5 logical drive
are defunct.
Three or more drives in a RAID level-5E logical drive are defunct.
If any of these are true, refer to "Rebuilding a defunct drive" on page 105.
The logical drive is okay. It is in a good, functional state.
The ServeRAID controller uses this reserved state during logical drive migration (LDM).
Chapter 3. Configuring controllers using the ServeRAID Manager program
39