Current Limitations - HP D7171A - NetServer - LPr Introduction Manual

white paper
Netserver NetServer NetServer
HP
Casualty Procedures
Failed FC Array Controller
In the event of an FC controller failure, the following steps should be taken:
Obtain a new controller and upgrade the firmware to the same version of firmware as the
working FC controller. If the firmware between the two controllers does not match, the
storage functionality will be disrupted and the cluster will fail. Perform the upgrade on a
separate server with an RS/12FC. Do not use a clustered RS/12FC.
Remove the failed controller from the RS/12FC.
Attach the cables to the new controller.
Insert the new controller. The light on the back of the new controller should blink yellow
for 1-5 minutes, then the light should turn green, and the RS/12 FC should return to
normal operation.
Verify controller's parameter settings using the FC Array Assistant.
Failed SCSI Cable / Controller
In the event of a failed SCSI cable or a failed SCSI controller, the following steps should be taken:
Obtain a new SCSI cable or controller.
Disengage the FC controller that is blinking red from the RS/12FC.
Replace the failed cable or insert the new controller. The light on the back of the new
controller should blink yellow for 1-5 minutes, then the light should turn green and RS/12 FC
should return to normal operation on it's own.
Verify controller's parameters setting using FC Array Assistant.

Current Limitations

One FC HBA Per Server
The main limitation of current SPOFless solutions is the lack of dual HBAs in the cluster nodes.
This is because the current FC driver does not support dual HBA functionality in case of HBA failure.
Because of this driver limitation there is still a single point of failure on each of the servers. However, the
failure of an HBA or link between the server and FC hub, or between the FC array controller and FC hub
will be considered as a cluster node failure and MSCS will take care of this situation by initiating a failover
to the surviving node.
Conclusion
The SPOFless solutions described in this document have passed all phases of the Microsoft
Hardware Compatibility Tests in configurations that meet all Microsoft
future, the HBA drivers are expected to support multiple FC loops so that more than one HBA
can be installed in the cluster servers, providing redundant paths to the storage.
solutions, as shown in the diagrams on page 5, will then become upgradeable, supporting a second HBA
in each server. Redundant HBAs will create a higher level of availability by representing a truly redundant
cluster. The diagram on the following page shows the SPOFless configuration meeting all redundant
requirements (redundant network adapters for public and private networks are not shown).
Documentation
®
requirements. In the
Current cluster
®
8