Disk file system showing as RAW

Software-based VM-centric and flash-friendly VM storage + free version

Moderators: anton (staff), art (staff), Max (staff), Anatoly (staff)

chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Wed Feb 14, 2018 12:18 am

Hi

Several times now I've had a REFS file system show as raw after a reboot. Is there a way of fixing the partition? I've found some utilities for rescuing the data, however all of these involve copying the data off and back to a new partition. A time consuming process.

What shutdown procedure should we be following to prevent any corruption?
I would like to add a script to the UPS shutdown procedure to safely shutdown star-wind. In the event of a power failure we let the servers run on battery for 5 minutes before triggering a shutdown. Ideally there would be two scripts, the first runs when power is lost, this would disable all write caching and create a snapshot. After 5 minutes the second script would shutdown starwind and the server

Any help is much appreciated.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Wed Feb 14, 2018 12:59 pm

Hi chell,
Could you please clarify, where are you using REFS, for StarWind or underlying storage?
ReFS is still a new file system and it can be "unstable". In your case, I would recommend using NTFS.
chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Wed Feb 14, 2018 1:27 pm

I have refs for the underlying storage and for the drive within starwind. The file system within starwind is the drive that has been corrupted.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Wed Feb 14, 2018 1:55 pm

Chell,
Can you please collect the logs from your systems and PM me for better understanding the problem you faced? Also, please specify more details about your system configuration.
You can collect log using this tool.
Thank you!
chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Wed Feb 14, 2018 9:46 pm

Hi

here is the log file https://www.dropbox.com/s/dh61uhx0mi4tz ... 1.zip?dl=0
This is a new two node set up running on Intel servers with windows 2016 standard. 128GB, 6x8TB in raid 6, 1TB SSD
I'm not sure when the corruption occurred. The server deployment started in December and then was put on hold due to another emergency. There's no data on the server yet , I would like to save the work done on the configuration. Above all I need to ensure it wont happen again.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Thu Feb 15, 2018 1:56 pm

Hi chell,
Thank you for logs. We will investigate the issue, it will take some time.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Thu Feb 15, 2018 2:29 pm

Chell,
Could you please provide logs from the second server as well?
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Fri Feb 16, 2018 9:02 am

Chell,
Thank you! We will proceed with the investigation.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Wed Feb 21, 2018 8:00 pm

Hi Chell,
Could you please follow these steps to avoid the issue with RAW disks:
1) Stop StarWindService. And do not forget to run the FlushCacheAll.ps1 script from StarWindX PowerShell examples folder before stopping the service;
2) rename .splumap and .spvmap extension in folders where failed devices located;
3) Start Starwind service;
4) Wait for the mounting progress;
Let me know how about the results!
chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Thu Feb 22, 2018 7:38 am

I've deleted the drive now.
I'll have to try that if it occurs again.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Thu Feb 22, 2018 10:34 am

Hi chell,
Thank you for your reply!
Ok, let me know if it occurs again.
chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Mon Feb 26, 2018 11:11 am

I've had a similar problem happen again. This time the drive was formatted as ntfs.
There was a power blackout. The UPS units triggered a shutdown but I cant be sure the shutdown completed before the battery power ran out. Shutting down starwind takes an incredible amount of time.
Now I can see a empty drive one one system, showing as offline (due to clustering)
The other system connects to the local iscsi. but no drive appears in drive manager.
I've tried the above suggestion and i have tried marking the drives as synced.
Oleg(staff)
Staff
Posts: 568
Joined: Fri Nov 24, 2017 7:52 am

Mon Feb 26, 2018 2:12 pm

Chell,
I will ask for logs from both servers again.
Could you please send the logs again?
We will investigate this issue.
chell
Posts: 48
Joined: Mon Dec 11, 2017 1:19 am

Mon Feb 26, 2018 3:05 pm

I'll generate them for you now.

Is it possible to roll back or mount the drive to an earlier version? (i assume that's what the spspx files are)
Post Reply