Sun Microsystems StorEdge 3000 Series, Инструкция по установке, эксплуатации и обслуживанию

Страница: 81 / 318

Chapter 5 First-Time Configuration 5-3
5.1.2 Dual-Controller Considerations
The following controller functions describe the redundant controller operation.
■
Both controllers must be exactly the same. They must operate with the same
firmware version, the same size of memory, and the same number of host and
drive channels. If one controller in a dual-controller configuration is replaced with
a field-replaceable unit (FRU) controller, the array automatically compares the
firmware versions of the two controllers. If they differ, the FRU controller
firmware is automatically changed to that of the existing controller in the array.
■
The controller firmware assumes that two controllers are available, or might be
made available during operation at any time. In a 1U single-controller
configuration, a 2U single-controller configuration, or a 2U dual-configuration,
once the primary controller (which might be the only controller) is powered on, it
begins to scan for a second controller. Until a second controller is discovered,
which does not happen in a 1U single-controller configuration or a 2U single-
controller configuration, the “Peripheral Device Status” for the redundant
controller will show a Status of “Scanning.” This is correct behavior and enables
the firmware to discover a second controller whenever it is added without the
necessity of rebooting the primary controller.
■
After booting in a redundant configuration, the controllers autonegotiate and
designate one controller as primary and the other controller as secondary.
■
The two controllers behave as one primary controller. Once the redundant
configuration takes effect, user configurations and settings can be done only on
the primary controller. The secondary controller then synchronizes with the
configuration of the primary controller, making the configurations of the two
controllers exactly the same.
The two controllers continuously monitor each other. When a controller detects
that the other controller is not responding, the working controller immediately
takes over and disables the failed controller.
■
It is necessary to connect all interfaces to both controllers so that the surviving
controller can readily continue all services provided for the RAID system. For
example, if you connect one controller to the Ethernet, you should also connect
the second controller to the Ethernet.
■
In an active-to-active configuration (standard configuration), you can assign any
logical drive to either of the controllers, and then map the logical configurations
to host channel IDs/LUNs. I/O requests from a host computer are directed to the
primary or the secondary controller accordingly. The total drive capacity can be
grouped into several logical configurations and assigned to both controllers so
that they share the workload. This active-to-active configuration engages all array
resources to actively maximize performance.
An active-to-standby configuration is also available but is not usually selected. By
assigning all the logical configurations of drives to one controller, the other
controller stays idle and becomes active only if its counterpart fails.