Version 1.0.0.440/441 and OneDrive Business Problems

[DriveDuck]

The creation of a new drive ends in error. “There was an error formatting a new drive. You can try to format it manually from Disk Management. Time out waiting for the new drive to arrive.”

 

Manually formatting the drive is not possible, the drive is offline. Trying to get the drive online also ends up in error “I/O Device error”.

 

I also noted that seconds after creating a new drive, the DrivePool Interface goes offline for about a minute.

[Christopher (Drashna)]

To make sure, check the disk number that StableBit CloudDrive lists in the UI (at the top) and which one is listed in Disk management. 

 

Additionally, if you're having issues with StableBit DrivePool at the same time, this may actually be related to "VDS", and may indicate a bigger issue with the system.

 

 

However, could you grab the logs from the system?

http://wiki.covecube.com/StableBit_CloudDrive_Log_Collection

[DriveDuck]

Used the disk number, of course. Don’t want to mess with the wrong drive. Always a good comment just to be sure.

 

It now looks as if upgrading from an older beta build did not go as smooth as it looked at first. Redid the 441 install and rebooted. After that I managed to successfully create a new drive with OneDrive for Business.

 

441 looks good now. I was a bit worried with all the changes Microsoft is doing with the new client they are currently launching. I would expect that there were or will be changes to the backend as well. No more 20.000 object limit as well as stuff like selective sync.

[DriveDuck]

Well, out of luck again. Again I can’t create a new drive. Rebooting did not make any difference this time.

 

VDS can’t be the problem since I don’t have it, the system in question is a Windows 2012R2 Datacenter Edition. DrivePool and CloudDrive are behaving normally otherwise.

 

DrivePool version is 2.1.1.561, should I upgrade to a newer version?

 

For privacy issues I need to simulate the setup on another system. If I’m able to recreate the problem I’ll send the logs.

[Christopher (Drashna)]

Well, OneDrive for Business "just" uses SharePoint as the backend, actually. 

 

As for the errors, could you enable logging, and reproduce the issue with CloudDrive.

http://wiki.covecube.com/StableBit_CloudDrive_Log_Collection

If you don't want to do the logging, could you grab the Error Reports from the system then? It's basically grabbing the files, and shouldn't contain any sensitive information. 

 

As for VDS, well, you definitely have it. It's a core component of Windows, and ALL installations have it. It shows up as "Virtual Disk" in the service list ("services.msc"), and it's what DISKPART and Disk Management connect to. 

 

As for StableBit DrivePool, upgrading to the Public Beta build may help, but i'm not sure.

 

 

 

And if you need to, then please to set up a separate VM. However, I have a sinking feeling that this may be a system specific issue.  I use HyperV to test on, and that includes just about every OS, including 2012R2 Standard/Datacenter, and I've not be able to reproduce the specific issue you mention. 

[DriveDuck]

Problem so far seems to be that DrivePool, when busy, causes the creation of a new drive with CloudDrive to take a long time. The drive creation formating runs into a timeout. Trying sometimes later sometimes works and I could create the drives. So far what I can see from the logs.

 

BTW Windwos 8 and higher as well as Windows Server 2012 and newer use Windows Storage Management, VDS has been superseded. Probably the API works very similar.

 

DrivePool has now been updated to the latest beta, I'll keep an eye if this makes any difference. Maybe it would be possible to extend the timeout during formating of a new CloudDrive?

[Christopher (Drashna)]

Well, if you can grab the logs, it should definitely help us.  And ideally the logs and error reports from both products. Just in case.

https://stablebit.com/Admin/IssueAnalysis/22875

 

 

As for Windows 8 and up, yup. Well sort of. For support of Storage Spaces, it uses a lower level of the system (WMI specifically, IIRC), but the VDS system and API still work exactly as expected. 

We are aware of that, but there isn't a reason to change StableBit DrivePool's code to reflect this, as we can't do anything with the underlying disks in a Storage Spaces pool anyways.

(and in case it matters, we are aware of it, and have noted the changes for StableBit Scanner, as this is absolutely relevant as it means we could display the status of drives in Storage Spaces). 

 

 

 

 

As for the timeout, I believe this is specifically provider dependant.  I'm not entirely sure, as it is one of the lower level parts.

However: http://wiki.covecube.com/StableBit_CloudDrive_Advanced_Settings

Try setting the "IoManager_HttpTimeoutMS" and the "IoManager_WriteRetries" values to higher numbers, restart the system, and see if that helps at all. 

[Christopher (Drashna)]

DriveDuck, was this still an issue for you?

 

If so, could you check something out for me?

Open an elevated command prompt (right click on it and select "run as admin").

Run "diskpart"

run "san" 

 

See what this is listed as. If it's anything other than "Online All",  that *could* be the cause of the issues, as the system may cause the drive to be offline when it's created. 

[DriveDuck]

The active policy is "OfflineShared", which is exactly what I would expect from the installed server version. Hope this helps.

[Christopher (Drashna)]

The active policy is "OfflineShared", which is exactly what I would expect from the installed server version. Hope this helps.

That's the normal for a server, it looks like.

 

And yeah, it means that any disk not on the same controller/bus as the system disk will be marked as "offline" by default. And that may definitely cause this issue. 

 

 

If you could run "san policy=onlineall" from diskpart, it should fix this, and prevent it from happening in the future. 

 

I've flagged the issue for Alex, so we can take a look at it and see about fixing the issue.

[DriveDuck]

I do get the drives online. Sometimes during the first attempt, sometimes I have to detach them a few times to get them online. Once they are online they stay online.

[Christopher (Drashna)]

Hmm, I've been having issues reproducing this issue, so it may be provider related.

 

If you could, enable drive tracing and then try reproducing the issue:

http://wiki.covecube.com/StableBit_CloudDrive_Log_Collection

[Christopher (Drashna)]

To clarify, the code does check to make sure the disk is online and writeable, regardless of the SAN policy (in fact, it's essentially "forced" online, if it's set to offline).

 

So, that you're seeing this indicates an issue with the provider, more than likely.  So please do get that logging, as it should indicate specifically where the issue is occurring.