Bug#827164: [Pkg-libvirt-maintainers] Bug#827164: unable to start guest with ceph rbd disks

2016-06-19 Thread uli iske
Am Fri, 17 Jun 2016 18:49:02 +0200
schrieb Guido Günther :

> On Wed, Jun 15, 2016 at 07:40:00AM +0200, uli iske wrote:
> > > On Tue, Jun 14, 2016 at 12:04:07PM +0200, uli iske wrote:
> > > [..snip..]  
> > > > Jun 14 11:51:32 kvm01 libvirtd[891]: Unable to read from
> > > > monitor: Die Verbindung wurde vom Kommunikationspartner
> > > > zurückgesetzt Jun 14 11:51:32 kvm01 libvirtd[891]: internal
> > > > error: early end of file from monitor, possible problem:
> > > > 2016-06-14T09:51:32.323392Z qemu-system-x86_64: -drive
> > > > file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
> > > > error connecting Jun 14 11:51:32 kvm01 libvirtd[891]: internal
> > > > error: process exited while connecting to monitor:
> > > > 2016-06-14T09:51:32.323392Z qemu-system-x86_64: -drive
> > > > file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
> > > > error connecting
> > > 
> > > This looks like an error connecting the Ceph disk so I'd either
> > > try to reproduce without libvirt or to leave out the disk from
> > > the domain XML and try to attach it via virsh
> > > qemu-monitor-command.
> > > 
> > > This might be related to the password-secret which is fairly new
> > > code but I'm just guessing here.
> > > 
> > > Tou could try to downgrade libvirt to 1.3.4 which did not have the
> > > secret code. If it works there we know where to look.  
> > 
> > You are right, I managed to install libvirt 1.3.3 and it works.  
> 
> …and QEMU is still the same version?

Yes, no change QEMU Version is still Version: 1:2.6+dfsg-1+b1

> Please attache the domain XML.


  sam
  db5a9297-e490-4013-b685-156cf4978ca2
  4194304
  4194304
  2
  
/machine
  
  
hvm

  
  




  
  
Nehalem
  
  



  
  destroy
  restart
  restart
  


  
  
/usr/bin/kvm

  
  

  
  



  
  
  
  
  
  


  
  


  
  
  


  
  
  


  
  
  


  
  


  
  


  


  
  
  
  
  
  


  
  
  
  
  
  


  
  
  


  
  
  


  
  
  


  


  


  


  
  
  


  


  


  
  
  

  


Regards

Uli
 
> Cheers,
>  -- Guido



Bug#827164: [Pkg-libvirt-maintainers] Bug#827164: unable to start guest with ceph rbd disks

2016-06-14 Thread uli iske
> On Tue, Jun 14, 2016 at 12:04:07PM +0200, uli iske wrote:
> [..snip..]
> > Jun 14 11:51:32 kvm01 libvirtd[891]: Unable to read from monitor:
> > Die Verbindung wurde vom Kommunikationspartner zurückgesetzt Jun 14
> > 11:51:32 kvm01 libvirtd[891]: internal error: early end of file
> > from monitor, possible problem: 2016-06-14T09:51:32.323392Z
> > qemu-system-x86_64: -drive
> > file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
> > error connecting Jun 14 11:51:32 kvm01 libvirtd[891]: internal
> > error: process exited while connecting to monitor:
> > 2016-06-14T09:51:32.323392Z qemu-system-x86_64: -drive
> > file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
> > error connecting  
> 
> This looks like an error connecting the Ceph disk so I'd either try to
> reproduce without libvirt or to leave out the disk from the domain XML
> and try to attach it via virsh qemu-monitor-command.
> 
> This might be related to the password-secret which is fairly new code
> but I'm just guessing here.
> 
> Tou could try to downgrade libvirt to 1.3.4 which did not have the
> secret code. If it works there we know where to look.

You are right, I managed to install libvirt 1.3.3 and it works.

Best Regards

Uli

> 
> Cheers,
>  -- Guido



Bug#827164: unable to start guest with ceph rbd disks

2016-06-13 Thread uli iske
I'll gladly be of assistance here, but I'm a bit lost
at this stage.

I have 2 hosts. All I can say so far is they both have the latest
Software Releases except qemu is 2.5+dfsg-5+b1 at the working host
and 2.6+dfsg-1+b1 at the non working host. 

libvirt-clients is 1.3.5-1 at both hosts.

This is what i get:

LIBVIRT_DEBUG=4 virsh start sam
error: Failed to start domain sam
error: internal error: process exited while connecting to monitor:
2016-06-13T14:16:07.875978Z qemu-system-x86_64: -drive
file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
error connecting

Working Host:

LIBVIRT_DEBUG=4 virsh start sam
Domain sam started

Regards

Uli


> Control: tags -1 -moreinfo
> Control: reassign -1 libvirt-clients
> 
> On Mon, Jun 13, 2016 at 01:34:02PM +0200, uli iske wrote:
> > Hmm, looks like qemu is not guilty here.
> > 
> > Guest is starting and accessing the ceph disk with the following
> > command:
> > 
> > qemu-system-x86_64 -drive
> > file=rbd:rbd/sam:id=admin:key=mysecretcephkey==:auth_supported=cephx:mon_host=10.210.0.5,format=raw
> > -machine pc-i440fx-2.4,accel=kvm -enable-kvm -curses
> > 
> > no errors. 
> > 
> > Sorry for the noise.  
> 
> Please hunt to original problem about the 'internal error: process
> exited while connecting to monitor'
> 
> Bugreport reassigned to libvirt-clients as my best guest
> 
> 
> Cheers
> Geert Stappers



Bug#827164: unable to start guest with ceph rbd disks

2016-06-13 Thread uli iske
Hmm, looks like qemu is not guilty here.

Guest is starting and accessing the ceph disk with the following
command:

qemu-system-x86_64 -drive
file=rbd:rbd/sam:id=admin:key=mysecretcephkey==:auth_supported=cephx:mon_host=10.210.0.5,format=raw
-machine pc-i440fx-2.4,accel=kvm -enable-kvm -curses

no errors. 

Sorry for the noise.

Regards

Uli

> 13.06.2016 12:31, uli iske wrote:
> 
> >> Please give complete qemu error message, not the interpretation
> >> given by libvirt.  
> > 
> > Is there a log somewhere or do I have to start the guest with
> > qemu-system-x86_64 ?  
> 
> I don't know libvirt enough to be able to answer this question, sorry.
> 
> Thanks,
> 
> /mjt



Bug#827164: unable to start guest with ceph rbd disks

2016-06-13 Thread uli iske
> Control: tag -1 + moreinfo
> 
> 13.06.2016 11:15, uli iske wrote:
> > Package: qemu
> > Version: 2.6+dfsg-1+b1
> > 
> > When trying to start or migrate Guest with ceph rbd Disk image
> > connected I get the following error message:
> > 
> > internal error: process exited while connecting to monitor:
> > 2016-06-13T08:11:05.686421Z qemu-system-x86_64: -drive
> > file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
> > error connecting  
> 
> Please give complete qemu error message, not the interpretation
> given by libvirt.

Is there a log somewhere or do I have to start the guest with
qemu-system-x86_64 ?

> Also please ensure you have qemu-block-extra of the right version.

qemu-block-extra is of Version 2.6+dfsg-1+b1
 
> Thanks,
> 
> /mjt



Bug#827164: unable to start guest with ceph rbd disks

2016-06-13 Thread uli iske
Package: qemu
Version: 2.6+dfsg-1+b1

When trying to start or migrate Guest with ceph rbd Disk image
connected I get the following error message:

internal error: process exited while connecting to monitor:
2016-06-13T08:11:05.686421Z qemu-system-x86_64: -drive
file=rbd:rbd/sam:id=admin:auth_supported=cephx\;none:mon_host=10.210.0.5\:6789\;10.210.0.6\:6789\;10.210.0.88\:6789,password-secret=virtio-disk0-secret0,format=raw,if=none,id=drive-virtio-disk0,cache=none:
error connecting

Same Guest works ok with qemu version 2.5+dfsg-5+b1

Regards

Uli



Bug#804781: vfs_snapper missing

2016-03-14 Thread uli iske
Hello,

works perfectly fine, Version is 4.3.6+dfsg-1

Thank you

Best Regards

Uli



Bug#813585: Cannot take external snapshot any more after active committing & pivoting

2016-02-03 Thread Uli Iske
Package: qemu
Version: 2.5+dfsg-4+b1

Pleae have a look at already reported and fixed BUG here:

https://bugzilla.redhat.com/show_bug.cgi?id=1300209

Same error message is with debian-testing, haven't tested with 2.5+dfsg-5 yet,
because fix is of date 2016-01-30 so I think it has not made it in.

Thanks and best regards

Uli



Bug#804781: vfs_snapper missing

2015-11-11 Thread Uli Iske
Package: samba-vfs-modules
Version: 4.3.0+dfsg-2

Architektur: amd64

Hello,

I'm testing Samba of ceph rbd. I'd like to test snapper Snapshots and
Windows previous versions with the new vfs_snapper module.
Could you please enable vfs_snapper module in your next builds ?

Tanky you and

Best regards

Uli


VFS_SNAPPER(8)System Administration tools
VFS_SNAPPER(8)

NAME
   vfs_snapper - Expose snapshots managed by snapper as shadow-copies

SYNOPSIS
   vfs objects = snapper

DESCRIPTION
   This VFS module is part of the samba(8) suite.

   The vfs_snapper VFS module exposes snapshots managed by snapper for use 
by Samba. This provides the ability for
remote
   SMB clients to access shadow-copies via Windows Explorer using the 
"previous versions" dialog.

   Snapshots can also be created and remove remotely, using the File Server 
Remote VSS Protocol (FSRVP). Snapshot
creation
   and deletion requests are forwarded to snapper via DBus.

   This module is stackable.

CONFIGURATION
   The underlying share path must have a corresponding snapper 
configuration file. The snapshot directory tree must
allow
   access for relavent users.

 [share]
 vfs objects = snapper

   For remote snapshot creation and deletion, Samba's FSRVP server must be 
configured in the [global] section:

 [global]
 rpc_daemon:fssd = fork
 registry shares = yes
 include = registry

PERMISSIONS
   Snapper stores snapshots under a .snapshots subdirectory. This directory 
must permit traversal for any users
wishing to
   access snapshots via the Windows Explorer previous versions dialog. By 
default, traversal is forbidden for all
non-root
   users. Additionally, users must be granted permission to list snapshots 
managed by snapper, via snapper's ALLOW_USERS
   or ALLOW_GROUPS options. Snapper can grant these users and groups 
.snapshots traversal access automatically via the
   SYNC_ACL option.

   Remote snapshot creation and deletion is only permitted by Samba for 
Active Directory administrators, backup
operators,
   or users explicitly granted SeBackupPrivilege. Snapper must also permit 
creation and deletion for the appropriate
user,
   via snapper's ALLOW_USERS or ALLOW_GROUPS options.

   The DiskShadow.exe FSRVP client initially authenticates as the Active 
Directory computer account. This account must
   therefore be granted the same permissions as the user account issuing 
the snapshot creation and deletion requests.

VERSION
   This man page is correct for version 4.3.0 of the Samba suite.

AUTHOR
   The original Samba software and related utilities were created by Andrew 
Tridgell. Samba is now developed by the
Samba
   Team as an Open Source project similar to the way the Linux kernel is 
developed.

Samba 4.3 09/28/2015
VFS_SNAPPER(8)