Sun Fire X2270 M2 Server
 
Sun Fire X2270 Server
 

PLA-Support>Sun Systems>Sun_Other>Sun Collections>SN-OTH: x64-CAP VCAP
 

In this Document

Goal

Solution

References

Applies to:

Goal

Verify the state of disks and/or internal RAID lun after service of the x2270/x2270M2

Solution

DISPATCH INSTRUCTIONS
WHAT SKILLS DOES THE FIELD ENGINEER/ADMINISTRATOR NEED: An onsite FE should be trained on the x2270 and x2270M2 server hardware.
TIME ESTIMATE: 30 minutes
TASK COMPLEXITY: 0

FIELD ENGINEER/ADMINISTRATOR INSTRUCTIONS:
- PROBLEM OVERVIEW: When the system board is replaced in these systems, the onsite FE should be able to configure the system board settings so that the new system board will access the disk drives as the defective system board had been doing. When a disk is replaced in an x2270 or x2270M2 server, the onsite FE should be able to check the status of the disks in the OS after the replacement. Information on how to manage the disk(s) when they are managed by different disk management environments is included in the Reference section below.
- WHAT STATE SHOULD THE SYSTEM BE IN TO BE READY TO PERFORM THE
RESOLUTION ACTIVITY?: Confirm that all hard disks are firmly seated in their bay sockets before booting the server.
- WHAT ACTION DOES THE FIELD ENGINEER/ADMINISTRATOR NEED TO TAKE:
These systems do not have a SAS disk HBA qualified to be used with them, and that means that the system board integrated SATA controller is the only certified disk controller available. There are no HBA's certified for use with external storage either.
Even though there are no PCI disk controller HBA's offered with these systems, there are still quite a few possible ways that the disk controller can be configured. Each system can have up to 4 SATA disks that are qualified and each has a Solid State Disk (SSD) qualified as well.
Even though the SATA controller is the only qualified disk controller, that does not mean that the systems have no hardware based RAID configurations either. These SATA controllers can be used with 2 disks of the same kind in a hardware RAID configuration, but at the same time this hardware RAID is only supported when a certified version of MS Windows is installed. Mixing of disk models in a volume is not supported, but a pair of disk bays can be used in a hardware RAID using the same model drives, while the other pair of disk bays can have a different model drive. See the reference links at the bottom of the document for instructions on what MS Windows versions can be installed, and what steps to take to get the MS Windows OS installed. While other x86 OS's may have SATA hardware RAID support, only SATA
hardware RAID with MS Windows is supported by Oracle since Intel, who makes the SATA controller, does not offer a driver for OS's other than MS Windows.

The first item of consideration after a disk (hard disk or SSD) is replaced in one of these systems is setting the BIOS as it used to be set in the replaced system board BIOS. In the BIOS, under the Advanced BIOS menu, is an IDE Configuration menu. In that menu the controller can be set to the legacy setting of IDE, a setting titled AHCI, and here is also were the system board has the SATA controller setting to enable hardware RAID. That is the extent of what the IDE Configuration menu is generally used for.
While a customer should already know what setting to use, here is some information on how the settings affect use of the connected disks. In IDE mode, the controller operates as has been the IDE standard for years. In IDE mode, the disk is available to the system with no further actions needing to be taken, and some OS's will not need an added driver to make use of the disk drive. When the controller is set to AHCI mode though, a performance boost is provided, and new functions (such has disk hot plug) become available also. In AHCI mode, some OS's will need a driver added during installation or the disk won't be accessible by that OS. The 3rd mode (RAID) is what enables use of the SATA controller hardware RAID functions. With the SATA controller set to RAID, the admin is expected to reboot the server and input cntl+i during BIOS boot to then configure the hardware RAID configuration itself. Again, the hardware RAID configuration is only supported with MS Windows OS installations. Also, the default BIOS setting in the IDE Configuration menu is AHCI.

When it comes to disk replacements (not a system board replacement), these BIOS settings should not need changing. As long as the controller is set to AHCI or RAID, a disk can also be hot plugged with no impact to the OS. If the disk is directly mounted by the OS (not under hardware based RAID), a disk that has not already been offlined by the OS must be manually unmounted by some OS's or the OS could crash if a disk is unplugged with power on.

If a system board is replaced, it is expected that the customer already knows what setting the BIOS IDE Configuration menu is set to. If the system admin is not aware of how the BIOS was set in the event of a system board replacement, it will be difficult to determine which IDE Configuration menu setting was previously used. In the event of system board replacement, if the system has an MS Windows OS is installed and hardware RAID was being used, when the RAID configuration is enabled the disk that is configured as the RAID mirror partner will be resynced and any data that is already on it will be overwitten from the mirror primary disk. If the system had hardware RAID configured before the system board was replaced, the OS might fail to boot if the hardware RAID configuration is not re-enabled.

A MOS document that includes information that is useful while diagnosing MS Widnows OS disk issues is MOS document 1011590.1. Although it has some dated material in it and does not mention the x2270 or x2270M2 servers, but it does offer web page links and information that might be useful while diagnosing MS Windows OS disk issues of various kinds.
The MOS document 1005530.1 provides a few Solaris OS commands that can be used when diagnosing disk issues in Solaris, though this document does not reference the x2270 or x2270M2 servers.
The MOS document 1002936.1 provides commands that may be helpful when diagnosing disk issues when a Linux OS is installed. Also, the MOS document 1008407.1 offers references to several other Oracle documents that could be helpful while trying to diagnose disk issues when working with a Linux OS.

OBTAIN CUSTOMER ACCEPTANCE
- WHAT ACTION DOES THE FIELD ENGINEER/ADMINISTRATOR NEED TO TAKE TO RETURN THE SYSTEM TO AN OPERATIONAL STATE: After the disk or system board action is complete, the Field Engineer should have the customer verify the status of the disks as well.

PARTS NOTE:

REFERENCE INFORMATION:
http://docs.oracle.com/cd/E19432-01/index.html (Sun Fire X2270 Server Documentation)
http://docs.oracle.com/cd/E19902-01/index.html (Sun Fire X2270 M2 Server Documentation)
<Document 1011590.1> How to check for Windows platform disk errors and online/offline status (Doc ID 1011590.1)
<Document 1005530.1> How to Check for Solaris[TM] X86 / X64 Disk Errors and Online/Offline Status (Doc ID 1005530.1)
<Document 1002936.1> How to Check for Linux Platform Disk Errors and Online/Offline Status (Doc ID 1002936.1)
<Document 1010946.1> General Guidance for Diagnosis and Replacing SCSI/SAS/SATA/FC-AL Disks running on Oracle Servers with Solaris (Doc ID 1010946.1)
<Document 1002753.1> How to replace a drive in Solaris[TM] ZFS (Doc ID 1002753.1)

References

This solution has no attachment