All HDDs go into unconfigured good after swapping one HDD - UCS C210

[yeopaul]

Hi all,

I have an issue with the UCS server that, after swapping the HDD and rebooted, it was fine for a while that everything is in good condition, however, the following day when viewing through CIMC, we have all the HDD showing as unconfigured good, and nothing was shown on the virtual disk info. Is the BIOS corrupted?

UCS C210 M2

Firmware version 1.4(3k)

BIOS version : c200.1.4.3f.0(Build Date:03/29/2012)

Controller : LSI 9261-8i MegaRaid SAS HBA

Appreciate your assistance.

regards,

paul

[Keny Perez]

Yeopal,

Something really important you left out was the RAID level you had configured, if you swapped HDDs and the RAID level was 5 your array is lost cause it only allows the failure of one disk; removing 2 disks is like losing 2 disks which RAID 5 will not support.

If the RAID level was 6, I would look to see if there is any other disk that might been on a "failed" state that could have been emulating the failure of the third disk.

Rate ALL helpful answers.

-Kenny

[uldericonucci]

Hi all,

I have exactly the same issue of yeopaul. Same server, same Controller, but BIOS 1.4.3k.0 and  CIMC 1.4(3v) (latest available).

The issue is active since I have upgraded the LSI RAID controller firmware, with HUU, to version 2.130.363-2183. The upgrade seems to be ok, no errors have been showed, but after rebooting the server all the Physical disks are in state "Unconfigured Good", and from CIMC I have no Virtual Drive showed under storage tab.

I'm not able to boot the OS (VMware), and on POST I have not the option to enable WEBBIOS (CTRL+H).

What can I do? Any help is really valued.

Thank you very much in advance,

Ulderico

[uldericonucci · Accepted solution]

Finally I have found a solution. I have hit the bug CSCuo07418. The problem is not related to RAID Controller but to Broadcom NetXtreme II BCM5709 4port NIC. The bug is still open but this workaround works:

“During POST, press Ctrl + S for launching the Broadcom OptionROM(this would be available only if the ROM is not disabled on the BIOS)

The main menu will show the available NIC ports

Enter each of the NIC card and perform the below operation

1. MBR Configuration

2. Boot Protocol

3. Set it to None”

Doing this procedure would let you boot the OS regulary. Hope this could helps others.

Regards,

Ulderico

[patrick.kofler]

This solution also helped me today to get an SNS-3415-K9 unit (which is actually an UCS C220M3 ) to detect its RAID controller.

Specs for those interested:

Firmware version: 2.0(8g)

BIOS version: C220M3.2.0.8.0 (Build Date: 07/16/2015)

Driver package containted in HUU iso ucs-c220-huu-2.0.8g.iso

[Cristian Munoz]

Hi Yeopal,

You can try to bring the configuration or the previous VD online by rebooting the server and select import configuration, If no all other hard drivers are failed or un-configured bad.

Foreign configuration(s) found on adapter.

Press any key to continue or `C' load the configuration utility,

or `F' to import foreign configuration(s) and continue.

Chris,  

[yeopaul]

Hi Kenny, Christian,

Thanks for replying.

It is a Raid 6, however only one HDD failed and being replaced. I will confirm to see if there is any additional HDD failure.

When it is showing all 10 HDDs are "unconfigured good" and not showing up on the virtual disk info, will that be the same on reboot, I just want to double confirm will loading configuration utility remove the Hosts configuration also?

regards,

yeopaul

[Cristian Munoz]

Hi Yeopaul,

When the server is power off, after a raid controller replacement or after moving HDDs to another server is commonly see the hard drives in un-configured good. After booting up the server, the raid controller will load the configuration and look for the existing VD but I haven't seen that after swapping just only one hard drive.

You can check the overall status of the hard drives by looking this outputs:

ucs-c2xx-m2 /chassis # show pci-adapter (to see the PCIe slot)

ucs-c2xx-m2 /chassis # scope storageadapter SLOT-X
ucs-c2xx-m2 /chassis/storageadapter # show virtual-drive

 
ucs-c2xx-m2 /chassis/storageadapter # show physical-drive
ucs-c2xx-m2 /chassis/storageadapter # show error-counters
ucs-c2xx-m2 /chassis/storageadapter # show physical-drive-count

Chris,

[yeopaul]

Hi Christian,

After confirming with the client that, all hosts on the UCS server are running fine.It is just not showing up on the CIMC and all drives appeared unconfigured good. Is there  a way to check if the controller if running fine? Could it be a CIMC issue and will rebooting CIMC helps? we are scheduling a time to do that.

paul

[Keny Perez]

Yeopal,

It is very hard to see that the host is up and running with no issues if all the disks are "unconfig good".  Is that server storing its data in any other place besides the local disks, like SAN for instance? That would be the only way...

I re-read this thread and it looks like we do not know what OS you are running, but if you install MegaCLI you will be able to run a few commands to check the controller's performance; commands like the ones listed here: 

https://supportforums.cisco.com/docs/DOC-16309

Now, there are some controllers that have a mechanism of "self-defense" and if you reseat a disk, the controller recognizes that the same disk that was there is in the same slot again and thinks "Well, if the disk was removed is for a reason, I don't know why it is here again, let's mark it as unconfig good" and that will force you to manually change that state and make it online manually.

Have you tried to reboot the server> press Ctrl+H when promted during POST and see if from Web BIOS the controller also says that the disks are in unconfig good?... This will confirm the real state of the disks.

Rate ALL helpfull answers

-Kenny

[jeffhuang]

I'm try to RESTART CIMC,everything is ok.