HPE EVA Storage

Ungrouped two failed FATA 1TB disk

 
Mathew Mengi
Occasional Advisor

Ungrouped two failed FATA 1TB disk

I have an EVA 4400 with three Disk Groups. One of the DG containing 8 FATA 1TB disks had two disks with the message "One loop connection lost".
The two disks in the group happened to loose one of their port connection at around the same time. From the eventlog it said that to ungroup the disk and then replace that. There is no orange light on the disk.

I would like to ungroup one of the disk and wait for the leveling process to complete and then add in the replacement disk and add that to the group and wait for the leveling process to complete. Once that is complete ungroup the next disk and repeat the above procedure.

What if the other disk looses both of its connection an failed when the leveling process is still running? Do a loose data on that Disk group? The current protection level is Single.

Someone can confirm or give a better way of solving that issue.

Ta
Mathew


4 REPLIES 4
Torsten.
Acclaimed Contributor

Re: Ungrouped two failed FATA 1TB disk

An advantage of a raid is you can loose disk and keep the data ...

If one loop connection was lost and is now back, you may consider to check controller and disk firmware.

Sometimes an update already fix the problem.

Hope this helps!
Regards
Torsten.

__________________________________________________
There are only 10 types of people in the world -
those who understand binary, and those who don't.

__________________________________________________
No support by private messages. Please ask the forum!

If you feel this was helpful please click the KUDOS! thumb below!   
Uwe Zessin
Honored Contributor

Re: Ungrouped two failed FATA 1TB disk

You cannot ungroup a disk from a disk group with 8 disk drives - period.
If you have to re-seat a disk, increase the failure detection interval before you pull the disk.

> The current protection level is Single.

The so-called "protection level" does not protect your data. Think of it as virtualized spare disks. The redundancy is in a virtual disk's VRAID-1, 5 or 6.

VRAID-0 does NOT offer any redundancy against disk drive failures - no matter how many objects on the EVA have the words "protection" or "redundant" in them.
.

Re: Ungrouped two failed FATA 1TB disk

Check the controller firmware and disk firmware. It is always recommended to make them in latest version.

Initially you can re-seat the drives within 10 seconds and check.

If possible can I request you to post the EVA logs (Controller Event Log and Configuration log) here.

I work for HPE
Accept or Kudo
hvhari
Esteemed Contributor

Re: Ungrouped two failed FATA 1TB disk

Hi, Levelling (a process for distributing data across all disks in the group to ensure more or less equal utilisation on all disks) and re-construction (a process to construct data from raid parity due to a failed disk) are different things. Levelling is a low priority back ground task & you can choose to add remove disks while this operation is in progress. However re-construction is a high priority job related to data availability & you should not be making any HW changes to array while this operation is in progress.

As others suggested, first thing to look here is firmware. If that is current you can look for replacing the disk after verifying the status (levelling or re-construction).
Regards,
Hari

If this post was useful , click the Kudos Star on the left side to say Thanks!