Re: [Nbd] NBD disconnects leading to squashfs error in LTSP

[Wouter Verhelst <w@...112...>]

On Sun, Aug 23, 2009 at 11:47:35AM +0300, Shahar Or wrote:
> Dear friends,
> 
> I'm using NBD in an LTSP setup.
> 
> I'm experiencing a problem where the following is printed in dmesg:
> 
> [   40.484917] nbd0: NBD_DISCONNECT
> [   40.486294] nbd0: Receive control failed (result -32)
> [   40.486585] nbd0: queue cleared
> 
> Following that are SQUASHFS errors, which I assume are a result of the
> NBD problem.

Indeed.

> The full dmesg log for the client is attached.
> 
> Is this a normal disconnect? What is result -32?

That's the error code that the server sent to the client at that point
in time. What it exactly means depends on the architecture of the
server, since errno is architecture-dependent.

> The server's daemon log shows at that time:
> 
> nbd_server[7925]: Read failed: Inappropriate ioctl for device

What are you serving to the client? Is that a regular file, or a raw
hard disk partition? Or perhaps something else?

-- 
The biometric identification system at the gates of the CIA headquarters
works because there's a guard with a large gun making sure no one is
trying to fool the system.
  http://www.schneier.com/blog/archives/2009/01/biometrics.html