Jump to content
Division-M Community
Requiem

BSOD with ntfs.sys caused by VHyperdrive5

Recommended Posts

Hi,

 

yesterday, my server once crashed, most likely due to a buggy plugin I was working with. I uninstalled the plugin and experienced crashes again and again, whe I saw that the server threw a BSOD With

"page fault in nonpaged area - ntfs.sys"

In the windows event log I found several errors I never experienced before:

 

"VHyperdrive5" EventID: 1

 

Error message was a number (varies), thus everytime preceeding the crash.

 

I restored several server backups (system drive) from a few days ago, sometimes the DB service would not start anymore, also tried 1.9.5 with v4 drivers with the same error.

Checkdisk of the system drive revealed no errors. Also no errors and BSODs when running Windows in Safe Mode.

 

I cannot remove disks from the pool or perform a pool repair because the system will crash before these tasks finish. I am running full checkdisk in safe mode on the pool's harddisk now

and I have raised a ticket, but no answer yet and google didn't help me either.

 

Maybe someone here knows an approach before I get reply from Anthony (who has all the logs and crashdump analysis along with the ticket).

 

Thanks for help!

Share this post


Link to post
Share on other sites

WHS2011, physical machine, not a VM.

 

This has always been running fine, as said, first crash came when fiddling with a plugin which I know is buggy. So I thought maybe a file system has been damaged or sth like that, but until now, the checks are running fine.

 

When in Windows safe mode, the machine keeps running and running and running without any problems, and no other tasks are throwing any errors, only DB. Overall, I don't think this is a hardware issue, but I'll run a memcheck. Any other ideas?

Share this post


Link to post
Share on other sites

Thanks a lot, anyway. I am using my time now to remove harddisks one by one and see if problems continue. At least the server crashed only 2 times last night (instead of crashing every 10 minutes or so). Though there are no errors reported by SMART ord by checkdisk, I am not finally sure that everything is okay with the harddrives. Too bad I can't pin it down....

Share this post


Link to post
Share on other sites

One of the first things I tried, yes. Unistall, reinstall, V1.9.5 as well as 2.3.0, as well as acomplete system restore from backup...

 

What I notice now: I am about to remove my two internal drives from the pool (all others are in an external USB3.0 case), and as the error persisted with only the two internal pooled drives running, I'd suppose that one of these might be faulty if it is a hardware issue (pool in fault tolerant mode).

Now, with the pool running with all harrdrives and removal of the two internal disks (chkdisk run fine on them with no errors, SMART is ok), I still keep getting The VHyperDrive5 log entries (but MUCH fewer, used to be there every few minutes) and my server hasn't crashed for almost 24 hours now.

This is a really, really annoying situation.

Share this post


Link to post
Share on other sites

Mmh, removed my internal harddrives from the pool and physically disconnected the drives - error messages and crashes stay, dumps are always ponting to the DB driver :-(

Hope I get an answer from Anthony soon after the weekend (and hope he can do something for me), this is really screwed up here.

Share this post


Link to post
Share on other sites

I will.

No crashes in the last 24 hours, Anthony got back to me and is investigating.

 

Oh, yeah, one "crash"... i dropped my external HDD with the server backups. I do not have my serverbackups now anymore :-\

Need to get a new backup disk.

Share this post


Link to post
Share on other sites

Windows Crashdumps and Event logs.

 

Crashes seemed to cease, I had none for several days.

But the I detected that I had multiple primary files in several directories. (I think this happened after a repair pool action which I tried because I thought maybe some corrupted pool information data might get repaired that is causing the crashes. DB told me that it recovered about 10000 missing primary files (they were NOT missing)).

Now it won't detect them as doubles, the file duplication tools won't remove the primary duplicates, my harddisks are running full so I had to disable duplication, and after I tried another pool repair - tadaa, my crashes are back, about every 10 minutes. This is really f*ed up beyond what I understand and I fear that I have to reinstall the complete server installation and install an whole new pool.

Damn :-(

Share this post


Link to post
Share on other sites

Okay, I think I have a solution.

I was about to setup my server from scratch yesterday and wanted to try one last thing as I noticed that the crashes were suddenly gone when my external Harddisk Bay was not connected and the pool was empty.

So, I uninstalled Drivebender, booted up the Server again, assigned drive letters to the external disks.

I ran Windows chkdsk on all of them (5 external disks), and on two of them chkdisk found and corrected errors (unfortunately it was not too specific about WHAT it has done...). Anyway, after this I removed the drive letters and installed DB again, using the latest driver updates from anthony after the installation finished and rebooted.

 

The Windows log entries of the driver are still there occasionally (but, according to Anthony, these are usually benign - anyway, I'd like to know what causes them to stop them from appearing. At last, this driver operates a lot of data).

 

I am still left with multiple primaries, but Anthony linked my to a Registry Value that can change the default behaviour and delete the multiples. It should do so on a file system health scan, but right now I have a queue of about 45000 files being validated.

DB came up with system alerts about multiple primaries before I changed the registry value to delete these files, I hope when Validation is finished, which may take some days, I'll be able to get rid of those automatically.

Share this post


Link to post
Share on other sites

Final feedback: Server running stable for a couple of days now. After a long file duplication queue which took a few days to process, the health scan identified all multiple primaries and deleted them automatically.

 

Still, the VHyperdrive5.sys errors keep appearing in the log files, apparently without any consequences..... wonder why.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...

×
×
  • Create New...