All Disks Failing in EVA 4400 Disk Enclosure

[ALSO]

Hi,

I am currently setting up our old EVA 4400 with 2 disk-enclosures for a labor environment. I uninitialised it and ugraded the management-module and the controllers to the newest available firmware.

Now i ran into the problem, that i can't initialise it because in Command View every single disk drive says:

"The disk drive in this bay is failed, not mated properly to the bay, or has lost power."

In addition i receive the following message when logging into the P6000 control panel:

"All disks groups in the storage system are RAID1 inoperable. Correct the situation and click Restart to continue operation." And at the beginning it asked if i want to continue loading the settings with only one quorum disk.

(see Screenshot)

I already tried to reseat all the disks and booted the whole system. It didn't change anything. Looking at the enclosure all disk-LEDs are green. The Disks I'm using are 300GB FC 15K and as far as I found out by googling, they should be compatible.

Could it be a problem with the disk drive firmware? If yes, there is no way to upgrade it without initialising then system right?

Thank you for your help.

[Jason_L]

Hello,

It appears that the EVA is trying to read metadata from the disks.  

Were any of the currently installed disks not part of this original system?

For instance, did some of the disks come from a different EVA?  Or perhaps were sitting on a shelf not installed until recently?

[PAC-MAN]

Hi,

I see a soft diagnostic failure in screenshots, During the BIST process if one of the test fails the controller will report soft diagnostic error, We can identify this from the event log check if there is any event reported for diagnostic failure, that event will tell us which component is failed.

Even though the failed hardware is recovered by the controller post reboot. controller will still have soft diagnostic error and it can be cleared only during BIST. I would recommend a restart for both controllers so that they will perform BIST again.

Thanks.

[Torsten.]

What firmware did you update (from, to)?

I would try to uninitialize again.

[ALSO]

Hello,

The disks used were installed in a EVA 6400 system and were not used for this EVA 4400.

I restarted both controllers now. I assume the software diagnostic failure is displayed because the cache battery of controller 2 is bad. After the reboot both controllers apear as failed in CommandView but only controller 2 has a software diagnostic failure in controlpanel.

The EVA is already uninitialised. I upgraded from 09534000 to 11001100

Thank you.

[Torsten.]

So the new controllers of the 4400 picked up parts of the 6400 config.

Try to initialize the array now.

[Jason_L]

If the disks were part of nother EVA and were not "ungrouped" prior to removal, then the metadata still resides on the disks.  When a controller powers up, it tries to read the metadata.

See the following advisory.

http://h20000.www2.hp.com/bizsupport/TechSupport/Document.jsp?objectID=c03664719&lang=en&cc=us&taskId=101&prodSeriesId=3664763&prodTypeId=12169

I assume there is no active data needed on the disks?

If there is no needed data on the disks, you can try allowing the EVA to come up on the one quorum disk it has detected (by answering yes) and as another suggested, try initializing the array again to recreate the proper quorum structure.

[ALSO]

Hello,

Now when I accept to load the metadate from that one quorum disk I recieve the other error mentioned:

All disks groups in the storage system are RAID1 inoperable. Correct the situation and click Restart to continue operation. So i clicked restart here and opend CommmandView.

All the disks are now still displayed failed in the CommandView. Due to that the EVA returns the message "There are not enough disks to meet the minimum number requirement." when trying to initialise, althought the disk-enclosures are fully populated.

Thank you for your help,

[Jason_L]

It appears there is still quorum confusion by the disks from the other EVA.  

I have been talking this over with some collegues and we agree on the next steps for this one.

You can try the following to remove the qurom disk from the EVA:

1. Power off the array

2. Then pull all the disks (except for 8 disks total) - which would be minimum configuration.

3. Start the EVA back up and see if it gives you the same quorum disk message

        - If it does, then you didn't pull the disk it identifies as the quorum, use process of elimination to remove it.

4. If it doesn't give the same message, try reinitializing the array again.

-Jason

[ALSO]

Hi Jason,

I followed your instructions and managed to eliminate the quorum disk. But now the EVA shows some really strange behaviour. The quorum disk message in the control-panel disappeared but when I now open Command View my components, like disk-enclosures or controllers, are not displayed anymore.

Therefore trying to reinitialise the array gives back the message: "Error getting list of physical disks.", and a refresh or discovery doesnt help.

Thank you.