Missing Datastore (EXSi 6.7)

[EvanBlank]

My Unity-VV-Users-Files and Unity-VV-VMs datastores have disappeared what should I do? I attached screens of the shell log of what I have tried so far along with other screenshots showing the errors and the site will not let me upload a support bundle (not sure how helpful that will be but I'm putting everything out there that I can).

[Alfista_PS]

Hi,

are they a local datastores or shared over network like NFS?

If they are from network I will suggest you to check the whole network if its OK and you can access the shared storages.

If it's local I will suggest you to come to the shell and first check the stores if they are OK and do you see it. 

Then You should try to unmount it and then again to mount.

And I have also fount this to your problem what you should also check:

 - This issue occurs if there is a mismatch between the Jumbo frame size in the router and the vmkernel MTU settings.

- https://kb.vmware.com/s/article/65169

Hope it helps.

[mguidini]

Hi @EvanBlank ,

If the datastores are provided from local disks, check the server management console for hardware errors and volume management.

In case they are NFS or iSCSI you will have vmkernels assigned to them, check for network connectivity to and from these vmkernel's IP Addresses.

I would also check what is the status of the volumes in the storage management interface. Assuming this is an EMC Unity (from the datastores name description), this interface is the Unity Unisphere, which can be accessible via browser. From there you will be able to check the status of your volumes and disks and if/how they are interacting with your vCenter/ESXi hosts. 

You will also be able to mount/unmount the volumes from there, as suggested.

Best,

[EvanBlank]

Everything looks fine in Unity. Do you see anything out of the oridinary? How do I check if/how they are interacting with your vCenter/ESXi hosts?

[Alfista_PS]

Have you check the network if all is OK and then the vmkernels?

If there is all OK then I will suggest you to check this not only in vCenter but on each ESXi host.

If there is also all OK then the next step should be to unmount all stores and mount it again or better you can fully remove it from each host and add it again as from scratch.

When this doesn't help then you have the last possibility to upgrade to the version 7 as its written in sent KB.

[mguidini]

There used to be a menu named Access>Hosts where you could find the list of ESXi hosts and their status, and some more info about the initiators, LUNs and IP Addresses.

[mguidini]

There used to be a menu named Access>Hosts where you could find the list of ESXi hosts and their status, and some more info about the initiators, LUNs and IP Addresses.

[RajeevVCP4]

is any outage on storage side 

did you check vmkernel.logs is any APD or PDL error triggered there, if yes then you need to reboot the host

[HailegebriealA]

If the storage is ISCSI you need to check the network and ISCSI IP configuration on vSphere or you can check on vCenter if any.

[EvanBlank]

How do I remount? ESXi doesn't like my NFS mount details. I might be doing it wrong

[EvanBlank]

Storage side looks good except Drive 23 and possibly a host initiator (not sure how to diagnose/fix that).

No APD or PDL errors in vmkernel log. I have attached the vmkernel log with all lines with the word "Error".

[EvanBlank]

Access>Hosts is missing 2 ip address and it won't let me add them. I can't add the initiator because it says host initiator already exists and host network address already exists

[Alfista_PS]

Hi,

this should help you unmount the broken filesystems:

https://kb.vmware.com/s/article/1005057

Its needed to do from shell while the WEB UI doesn't do it while the mounts are broken and working correctly.