smama21
Contributor
Contributor

RAID 1 is lost and inaccessible guests after activating the two hard drives

Greetings ALL,

I have the below issue and appreciate your support.

I have DELL PowerEdge R710 with two hard drive used to be RAID 1. VMware host is serving two VM guests.

With some power supply issues we encountered, I found the server is not working. Checked the BIOS and found the two drives are INACTIVE. I proceeded and activated them through SAS Controller (CTRL-C). However, after activation, I noticed the RAID-1 is not anymore applicable. That is drives are not member of RAID anymore unfortunately.

I proceeded with loading the host. VMware loading but reached to an error indicating duplicate UUID. I understood is related to the two drives (they have same ID - could be) since they are mirroring to each other (my understanding only).

I pulled the second drive physically and load the host again. VMware loaded normally. However when I remote connect to it using vSphere, I saw the two guests showing inaccessible (Unknown). I noticed there is no datastore.

I have images of the two VM guests (vmdk)but not I do not have image for the whole two drives when RAID 1 was working.

Two kind requests for help. I am new to VMware

One which is to understand is the VMware installed in the drives or with the BIOS.

Second, How I can return RAID1 and the VM guests. If not both, how to retrieve the VM guests using single drive.

Help please.

Regards,

Saeed

smama21

6 Replies
IRIX201110141
Virtuoso
Virtuoso

Please run "esxcfg-volume -l" from the command line and report the output.

Regards

Joerg

smama21
Contributor
Contributor

where would I run this command?

0 Kudos
ThompsG
Virtuoso
Virtuoso

Hi smama21,

You run the command from the ESXi console either through a PuTTY session or directly at the console. I would go the PuTTY route as it is easier to capture the output - here is a KB article on what to do to enable access: https://pubs.vmware.com/vsphere-6-5/index.jsp?topic=%2Fcom.vmware.vcli.getstart.doc%2FGUID-C3A44A30-...

Hope this helps.

smama21
Contributor
Contributor

Hi,

Here is the output:

~ # esxcfg-volume -l

VMFS UUID/label: 50c4bee4-b6a8cb8e-192d-d4ae52c0525c/datastore1

Can mount: Yes

Can resignature: Yes

Extent name: naa.5000c50041d218bf:3     range: 0 - 471295 (MB)

0 Kudos
ThompsG
Virtuoso
Virtuoso

Hi smama21​,

As you have mentioned, it appears your mirror (RAID1) has broken with the two drives being seen as individual drives but exact copies of each other. As you have removed one of the drives, you should be able to mount the datastore with the following command:

esxcfg-volume -M 50c4bee4-b6a8cb8e-192d-d4ae52c0525c

The "-M" makes the mount persistent otherwise you will lose access again after a restart.

In regards to recreating the RAID set again this is something that would need to happen with the RAID controller however I suggest you gain access to your VMs first and then ensure you have a good backup/recovery position before continuing.

Kind regards.

smama21
Contributor
Contributor

Thanks the VM guests returned to work again.

Thank you alot

I don't know how to thank you and the colleague responded above.

As you recommended, I will take backups. That is VM backups and drive image backups as well. Then I will sort out how to return RAID 1. Usually, I understood drives have to be formatted in order to enable RAID.

0 Kudos