[Virtio-fs] Getting help setting up virtiofs guest

[Virtio-fs] Getting help setting up virtiofs guest

[David Lomas]

[-- Attachment #1: Type: text/plain, Size: 1341 bytes --]

Apologies if this isn't the correct place for this kind or request.

I've been following these instructions to set up host and guest:
https://virtio-fs.gitlab.io/howto-qemu.html and
https://libvirt.org/kbase/virtiofs.html

On the host, I've built the latest version of libvirt (6.4.0) and qemu
(5.0.0), and added sections to the domain definition. All appears well (no
errors at least). This is the relevant section:

    <filesystem type='mount' accessmode='passthrough'>
      <driver type='virtiofs'/>
      <binary path='/usr/libexec/virtiofsd'/>
      <source dir='/fast_data/test'/>
      <target dir='test'/>
      <address type='pci' domain='0x0000' bus='0x00' slot='0x06'
function='0x0'/>
    </filesystem>

On the guest, I've build the kernel (4.19.28), booted it, and confirmed
virtio_fs.ko is a built-in module.

But on running mount -t virtio_fs test /mnt/test/ I get the following error:

mount: /mnt/test: wrong fs type, bad option, bad superblock on qube-server,
missing codepage or helper program, or other error.

Note both those web pages above suggest using mount -t virtiofs, but that
gives unknown filesystem type 'virtiofs'

I'm at a loss to find more detailed information on how to configure
this—are there other resources online, or can anyone suggest what to try
next?

Thanks!

[-- Attachment #2: Type: text/html, Size: 1900 bytes --]

Re: [Virtio-fs] Getting help setting up virtiofs guest

[Dr. David Alan Gilbert]

* David Lomas (dl3@pale-eds.co.uk) wrote:
> Apologies if this isn't the correct place for this kind or request.

It's the right place to ask.

> I've been following these instructions to set up host and guest:
> https://virtio-fs.gitlab.io/howto-qemu.html and
> https://libvirt.org/kbase/virtiofs.html
> 
> On the host, I've built the latest version of libvirt (6.4.0) and qemu
> (5.0.0), and added sections to the domain definition. All appears well (no
> errors at least). This is the relevant section:
> 
>     <filesystem type='mount' accessmode='passthrough'>
>       <driver type='virtiofs'/>
>       <binary path='/usr/libexec/virtiofsd'/>
>       <source dir='/fast_data/test'/>
>       <target dir='test'/>
>       <address type='pci' domain='0x0000' bus='0x00' slot='0x06'
> function='0x0'/>
>     </filesystem>

I think that's OK.

> On the guest, I've build the kernel (4.19.28), booted it, and confirmed
> virtio_fs.ko is a built-in module.

Where did you get that kernel?  It sounds rather old;  the main part of
virtiofs is now in the upstream kernel; so if your distro has a recent
5.x kernel, you can just use it.

> But on running mount -t virtio_fs test /mnt/test/ I get the following error:
> 
> mount: /mnt/test: wrong fs type, bad option, bad superblock on qube-server,
> missing codepage or helper program, or other error.
> 
> Note both those web pages above suggest using mount -t virtiofs, but that
> gives unknown filesystem type 'virtiofs'

It changed in the newer kernels to the 'virtiofs'.

> I'm at a loss to find more detailed information on how to configure
> this—are there other resources online, or can anyone suggest what to try
> next?

I suspect you're suffering from the older kernel here.  Try a nice
shiny new one.

Dave

> Thanks!

> _______________________________________________
> Virtio-fs mailing list
> Virtio-fs@redhat.com
> https://www.redhat.com/mailman/listinfo/virtio-fs

--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [Virtio-fs] Getting help setting up virtiofs guest

[David Lomas]

[-- Attachment #1: Type: text/plain, Size: 2574 bytes --]

Just to say thanks to Dave (all the best people are called Dave) for the
pointer to use kernel 5.7. That just worked on the guest, and I can now see
the host filesystem.

Perhaps this page <https://virtio-fs.gitlab.io/howto-qemu.html> should just
link to the mainline kernel rather than the virtio-fs one?



On Thu, Jun 11, 2020 at 8:11 PM Dr. David Alan Gilbert <dgilbert@redhat.com>
wrote:

> * David Lomas (dl3@pale-eds.co.uk) wrote:
> > Apologies if this isn't the correct place for this kind or request.
>
> It's the right place to ask.
>
> > I've been following these instructions to set up host and guest:
> > https://virtio-fs.gitlab.io/howto-qemu.html and
> > https://libvirt.org/kbase/virtiofs.html
> >
> > On the host, I've built the latest version of libvirt (6.4.0) and qemu
> > (5.0.0), and added sections to the domain definition. All appears well
> (no
> > errors at least). This is the relevant section:
> >
> >     <filesystem type='mount' accessmode='passthrough'>
> >       <driver type='virtiofs'/>
> >       <binary path='/usr/libexec/virtiofsd'/>
> >       <source dir='/fast_data/test'/>
> >       <target dir='test'/>
> >       <address type='pci' domain='0x0000' bus='0x00' slot='0x06'
> > function='0x0'/>
> >     </filesystem>
>
> I think that's OK.
>
> > On the guest, I've build the kernel (4.19.28), booted it, and confirmed
> > virtio_fs.ko is a built-in module.
>
> Where did you get that kernel?  It sounds rather old;  the main part of
> virtiofs is now in the upstream kernel; so if your distro has a recent
> 5.x kernel, you can just use it.
>
> > But on running mount -t virtio_fs test /mnt/test/ I get the following
> error:
> >
> > mount: /mnt/test: wrong fs type, bad option, bad superblock on
> qube-server,
> > missing codepage or helper program, or other error.
> >
> > Note both those web pages above suggest using mount -t virtiofs, but that
> > gives unknown filesystem type 'virtiofs'
>
> It changed in the newer kernels to the 'virtiofs'.
>
> > I'm at a loss to find more detailed information on how to configure
> > this—are there other resources online, or can anyone suggest what to try
> > next?
>
> I suspect you're suffering from the older kernel here.  Try a nice
> shiny new one.
>
> Dave
>
> > Thanks!
>
> > _______________________________________________
> > Virtio-fs mailing list
> > Virtio-fs@redhat.com
> > https://www.redhat.com/mailman/listinfo/virtio-fs
>
> --
> Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
>
>

[-- Attachment #2: Type: text/html, Size: 3829 bytes --]

Re: [Virtio-fs] Getting help setting up virtiofs guest

[Dr. David Alan Gilbert]

* David Lomas (dl3@pale-eds.co.uk) wrote:
> Just to say thanks to Dave (all the best people are called Dave) for the
> pointer to use kernel 5.7. That just worked on the guest, and I can now see
> the host filesystem.
> 
> Perhaps this page <https://virtio-fs.gitlab.io/howto-qemu.html> should just
> link to the mainline kernel rather than the virtio-fs one?

Oops yes! We need to update to say you can either use upstream, or use
the virtio-fs-dev branch of that git.

Dave

> 
> 
> On Thu, Jun 11, 2020 at 8:11 PM Dr. David Alan Gilbert <dgilbert@redhat.com>
> wrote:
> 
> > * David Lomas (dl3@pale-eds.co.uk) wrote:
> > > Apologies if this isn't the correct place for this kind or request.
> >
> > It's the right place to ask.
> >
> > > I've been following these instructions to set up host and guest:
> > > https://virtio-fs.gitlab.io/howto-qemu.html and
> > > https://libvirt.org/kbase/virtiofs.html
> > >
> > > On the host, I've built the latest version of libvirt (6.4.0) and qemu
> > > (5.0.0), and added sections to the domain definition. All appears well
> > (no
> > > errors at least). This is the relevant section:
> > >
> > >     <filesystem type='mount' accessmode='passthrough'>
> > >       <driver type='virtiofs'/>
> > >       <binary path='/usr/libexec/virtiofsd'/>
> > >       <source dir='/fast_data/test'/>
> > >       <target dir='test'/>
> > >       <address type='pci' domain='0x0000' bus='0x00' slot='0x06'
> > > function='0x0'/>
> > >     </filesystem>
> >
> > I think that's OK.
> >
> > > On the guest, I've build the kernel (4.19.28), booted it, and confirmed
> > > virtio_fs.ko is a built-in module.
> >
> > Where did you get that kernel?  It sounds rather old;  the main part of
> > virtiofs is now in the upstream kernel; so if your distro has a recent
> > 5.x kernel, you can just use it.
> >
> > > But on running mount -t virtio_fs test /mnt/test/ I get the following
> > error:
> > >
> > > mount: /mnt/test: wrong fs type, bad option, bad superblock on
> > qube-server,
> > > missing codepage or helper program, or other error.
> > >
> > > Note both those web pages above suggest using mount -t virtiofs, but that
> > > gives unknown filesystem type 'virtiofs'
> >
> > It changed in the newer kernels to the 'virtiofs'.
> >
> > > I'm at a loss to find more detailed information on how to configure
> > > this—are there other resources online, or can anyone suggest what to try
> > > next?
> >
> > I suspect you're suffering from the older kernel here.  Try a nice
> > shiny new one.
> >
> > Dave
> >
> > > Thanks!
> >
> > > _______________________________________________
> > > Virtio-fs mailing list
> > > Virtio-fs@redhat.com
> > > https://www.redhat.com/mailman/listinfo/virtio-fs
> >
> > --
> > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
> >
> >

> _______________________________________________
> Virtio-fs mailing list
> Virtio-fs@redhat.com
> https://www.redhat.com/mailman/listinfo/virtio-fs

--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK

Re: [Virtio-fs] Getting help setting up virtiofs guest

[Dr. David Alan Gilbert]

* Dr. David Alan Gilbert (dgilbert@redhat.com) wrote:
> * David Lomas (dl3@pale-eds.co.uk) wrote:
> > Just to say thanks to Dave (all the best people are called Dave) for the
> > pointer to use kernel 5.7. That just worked on the guest, and I can now see
> > the host filesystem.
> > 
> > Perhaps this page <https://virtio-fs.gitlab.io/howto-qemu.html> should just
> > link to the mainline kernel rather than the virtio-fs one?
> 
> Oops yes! We need to update to say you can either use upstream, or use
> the virtio-fs-dev branch of that git.

Fixed.

> Dave
> 
> > 
> > 
> > On Thu, Jun 11, 2020 at 8:11 PM Dr. David Alan Gilbert <dgilbert@redhat.com>
> > wrote:
> > 
> > > * David Lomas (dl3@pale-eds.co.uk) wrote:
> > > > Apologies if this isn't the correct place for this kind or request.
> > >
> > > It's the right place to ask.
> > >
> > > > I've been following these instructions to set up host and guest:
> > > > https://virtio-fs.gitlab.io/howto-qemu.html and
> > > > https://libvirt.org/kbase/virtiofs.html
> > > >
> > > > On the host, I've built the latest version of libvirt (6.4.0) and qemu
> > > > (5.0.0), and added sections to the domain definition. All appears well
> > > (no
> > > > errors at least). This is the relevant section:
> > > >
> > > >     <filesystem type='mount' accessmode='passthrough'>
> > > >       <driver type='virtiofs'/>
> > > >       <binary path='/usr/libexec/virtiofsd'/>
> > > >       <source dir='/fast_data/test'/>
> > > >       <target dir='test'/>
> > > >       <address type='pci' domain='0x0000' bus='0x00' slot='0x06'
> > > > function='0x0'/>
> > > >     </filesystem>
> > >
> > > I think that's OK.
> > >
> > > > On the guest, I've build the kernel (4.19.28), booted it, and confirmed
> > > > virtio_fs.ko is a built-in module.
> > >
> > > Where did you get that kernel?  It sounds rather old;  the main part of
> > > virtiofs is now in the upstream kernel; so if your distro has a recent
> > > 5.x kernel, you can just use it.
> > >
> > > > But on running mount -t virtio_fs test /mnt/test/ I get the following
> > > error:
> > > >
> > > > mount: /mnt/test: wrong fs type, bad option, bad superblock on
> > > qube-server,
> > > > missing codepage or helper program, or other error.
> > > >
> > > > Note both those web pages above suggest using mount -t virtiofs, but that
> > > > gives unknown filesystem type 'virtiofs'
> > >
> > > It changed in the newer kernels to the 'virtiofs'.
> > >
> > > > I'm at a loss to find more detailed information on how to configure
> > > > this—are there other resources online, or can anyone suggest what to try
> > > > next?
> > >
> > > I suspect you're suffering from the older kernel here.  Try a nice
> > > shiny new one.
> > >
> > > Dave
> > >
> > > > Thanks!
> > >
> > > > _______________________________________________
> > > > Virtio-fs mailing list
> > > > Virtio-fs@redhat.com
> > > > https://www.redhat.com/mailman/listinfo/virtio-fs
> > >
> > > --
> > > Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
> > >
> > >
> 
> > _______________________________________________
> > Virtio-fs mailing list
> > Virtio-fs@redhat.com
> > https://www.redhat.com/mailman/listinfo/virtio-fs
> 
> --
> Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK
> 
> _______________________________________________
> Virtio-fs mailing list
> Virtio-fs@redhat.com
> https://www.redhat.com/mailman/listinfo/virtio-fs
--
Dr. David Alan Gilbert / dgilbert@redhat.com / Manchester, UK