SUN 6540 一块硬盘故障，新购一块替换之后发现状态异常，磁盘识别为未认证状态，以下回答来自oracle 论坛，仅供各位参考使用，另外说一下百度搜索实在是惨不忍睹，除了广告没有什么有价值的信息。

Storage 6140 disk replacement because of DacStore Revision

 

 Salam all,
We have a Storage Tek 6140 running firmware 6.XX and managed by CAM, we recently bought many new disks for support purpose, when we tried to replace a failed disk the Storage didn't recognize the new disk and log an alarm in the CAM similar to this :     

    Severity:     Critical
    Date:     <date>
    State:     Open
    Acknowledged By:
    Auto Clear     Yes
    Description:     An uncertified drive has been detected: Tray.XX.Drive.XX
    Info:
    Device:     <array-name>
    Component:     Tray.XX.Drive.XX
    Event Code:     57.66.1027
    Aggregated Count:     <count

Our search revealed that the disks we bought has an existing DacStore on them (Normally new disk come with no DacStore), the DacStore revision is 4.
Firmware 6.XX deal with DacStore Rev 3 and firmware 7.XX deal with DacStore Rev 4, if you insert a disk with DacStore Rev 3 on a Storage running firmware 7.XX the Storage will upgrade automatically the DacStore to Rev 4 and the disk will work fine, this not the case if you insert a disk with DacStore Rev 4 into a Storage running firmware 6.XX, in this case the disk will not be recognized and the perevious Alarm will show up in the CAM.

Two possible solution, either we upgrade the Storage firmware to 7.XX which we rather prefer to avoid or we can simply clear the DacStore.

On Oracle Bug 6746964 document they are saying :

Access the serial shell and clear the DACstore.

Shell is not a user friendly environment and can make matters worse if the
wrong drive is cleared.

So if any one have the exact procedure for that or have another solution your help will be most welcome

Regards.

 

像这种情况，一定要注意控制器版本，同时也不建议进入命令行去操作，毕竟数据无价，避免不值得的操作。

本例情况就是和供应商重新换一块就好了。

 

附：请勿随意操作！

Controller firmware 6.xx and earlier have a different metadata (DACstore) data structure than controller firmware 7.xx and later. Metadata consists of the array and logical drive configuration data for a specific DS4000 or DS5000 storage subsystem. These two metadata data structures are not interchangeable. When powered up and in optimal operational state, the storage subsystem with controller firmware version 7.xx and later will be able to convert the metadata from the drives configured in storage subsystems with controller firmware version 6.xx and earlier to controller firmware version 7.xx metadata data structure. However, the storage subsystem with controller firmware version 6.xx and earlier cannot read the metadata from the drives configured in storage subsystems with controller firmware version 7.xx and later.

shellUsr

wy3oo&w4

 

1) unplug all drives

2) connect to the serial or via rlogin to one controller (doesn't matter

which one)

3) insert a drive

-> you get an output like this:

04/01/08-11:13:27 (GMT) (ssmDriveSpinTask): NOTE:

cfgCommonReplaceCode: dev = 0x400002

04/01/08-11:13:29 (GMT) (ssmDriveSpinTask): WARN: Unsupported

dacstore: dev = 0x400002

(ssmDriveSpinTask): NOTE: cfgCommonReplaceCode: complete dev =

0x400002

4) note down the devnum (0x400002 in this case)

5) rewrite the dacstore area: isp cfgWipe1,<devnum>

(isp cfgWipe1,0x400002 for this example)

6) repeat steps 3 to 5 for all drives

7) if done, reboot bot controllers and you will see all drives

Case closed, I appreciated your help Alexander, thank you.