Sudden Power Outage now Corrupt files

  • 10 January 2018
  • 2 replies

Badge +1
Ok Everyone,

I am fairly new to the Nutanix system as our company recently decided to upgrade all server hardware to a Hyper-Converged system. Because of this, my knowledge of the backupsnapshot systems used by Nutanix is very limited, so please bear with me.

So heres the issue; yesterday afternoon one of UPS burned itself up and caused a sudden power outage in our core datacenter. This location houses our primary Nutanix 4 node cluster. This is also the location where all employee profiles are stored using the Nutanix File Server system. After power was restored and th Nutanix system powered back up (took roughly 30mins) all services seemed to be restored. The next morning however, we started recieving calls about users not being able to login to the system. They are getting errors about network profiles not being available. Quicking checking the HomeShare location, we quickly determined the users home profile folder is suffering some issues. Properties shows 0bytes and all tabs are essentially blank. This leads me to believe the folder is corrupt and needs to be restored. Seems like a simple task with the exception that I have never done this before on a Nutanix system.

When i go into the Data Protection section, I can see that there is a local Snapshot of the File Server in question from yesterday at 9:45am (before the unexpected shutdown). My question now becomes, what is the process of restoring the system to this snapshot, or in the very least gaining access to it so i can pull the necessary files from it?

This topic has been closed for comments

2 replies

Userlevel 2
Badge +19
Did you have Nutanix guest tools installed on the VM? If so, click on the Nutanix guest tools icon on the desktop, login and use it to mount the snapshot on a drive letter and from there you can explore and copy files.

If not, restore the snapshot as a different VM and set a new IP address on it so that you can copy files.

Here's a good blog post.
Badge +1
Alright, my issue has been resolved. Here is what happened.

Since we are using the AFS system for the File Server, a virtual server is created with the name: D02-V02-001
This file server has 3 vms created for it with the following names: NTNX-D02-V02-001-1, NTNX-D02-V02-001-2 and NTNX-D02-V02-001-3
After verifying the DFS properties on the profile folders in question it seems all users with profiles on the NTNX-D02-V02-001-1 server were the unreachable folder. Rebooted the NTNX-D02-V01-001-1 vm and the previously troublesome folders are now accessible. Properties on the DFS tab shows the folder as good.

As a side note, using the console from Prism to the server in question (NTNX-D02-V02-001-1) had an error: Dangling Symlink, which i believe was the issue. There might be another command that will repair all "dangling symlinks", but the reboot was what fixed it for me.

Hope this helps someone else in the future.