CHAPTER 6 Redundancy
6.4 Redundancy of the Management LAN
100
Redundancy procedure Single volume Inter-volume Boot path
cabinet:
External RAID
device
Redundancy can be achieved in a single
volume, since the hardware RAID configuration
is used in the external RAID device.
Redundancy of boot paths is also possible by
installing a multi-
path configuration. Redundancy
of volume can also be achieved by configuring
software mirroring between volumes of multiple
RAID cabinet (using PRIMECLUSTER GDS). If
a problem occurs in the cabinet on one side, it is
possible to continue the system boot and system
startup in the cabinet on the other side.
is possible
is possible
is possible
6.4 Redundancy of the Management LAN
The Management LAN is used for connecting the MMB Web-UI and SVmco. The management LAN of each
partition is connected using a LAN port of IOU or a LAN card added in PCI Express slot to which the partition
belongs. This management LAN is connected through MMB user support. The MMB has four external
interface ports.
Two ports are allocated for maintenance, and the other two ports are used for connecting the console units
or connecting nodes between cabinet in the cluster configuration, to create duplex configuration architecture
(duplication).
For details on the management LAN, see '1.3 Management LAN ' in the “PRIMEQUEST 2000 Series
Administration Manual” (CA92344-0537).
6.5 Operation LAN Redundancy
This section describes the operational LAN redundancy that applies to the configuration of the operation
system. Configuration of a redundant network or and network interface requires at least two network
interfaces and an optional software for switching these interfaces.
The operation LAN redundancy supported by the PRIMEQUEST 2000 series is as follows.
- Duplication of a transmission path between servers (high-speed switching method)
Depending on the duplex network configuration of the transmission path due to PRIMECLUSTER GLS
(Global Link Services) or the type of communication partner server, the transmission path can be
duplicated by various procedures to achieve high reliability of the entire communication.
For details on the redundancy of operation LAN based on PRIMECLUSTER GLS, see PRIMECLUSTER
Global Link Services Manual (Transmission path duplex functions edition).
- Teaming with Intel PROSet
It is possible to configure a teaming based on Intel PROSet. For more information, see Help in Intel
PROSet.
- Redundancy using bonding
Linux supports redundancy using bonding.
- Redundancy using NIC teaming function
Windows Server 2012/ 2012R2 support redundancy using NIC teaming function that is provided by the
operating system.
Note
There are some precautions on teaming with Intel PROSet ®. For information on the items, see 'G.9 NIC
(Network interface card)' of the “PRIMEQUEST 2000 Series Administration Manual” (CA92344-0537).
6.6 Degradation Function
The degradation function disconnects (degradation) a faulty component when there is a hardware failure. To
continue operation, it configures the system with the remaining hardware resources. The system may be
restarted when using the degradation function.
System operation can be resumed immediately without a prolonged system failure because the system is
started with the faulty component disconnected.
To Next Page
Loading...
